Benchmarking Knowledge-assisted Kriging for Automated Spatial Interpolation of Wind Measurements
|
|
- Allyson Greene
- 5 years ago
- Views:
Transcription
1 Benchmarkng Knowledge-asssted Krgng for Automated Spatal Interpolaton of Wnd Measurements Zlatko Zlatev, Stuart E. Mddleton, Galna Veres IT Innovaton Centre Unversty of Southampton Southampton, UK Abstract - We have benchmarked a novel knowledgeasssted krgng algorthm that allows regons of spatal coheson to be specfed and varograms calculated for each regon. The varogram calculaton tself s automated and spatal regons created va offlne automated segmentaton of ether expert-drawn Google Earth polygons or ASA alttude data. Our use-case s to create nterpolated wnd maps for nput nto a bathng water qualty model of mcrobal contamnaton. We benchmark our knowledge-asssted krgng algorthm aganst 7 other algorthms on UK met-offce wnd data (89 sensors). Our wnd estmaton results are comparable to standard ordnary krgng usng varograms created by an expert. When usng spatal segmentaton we fnd our krgng error maps reflect better the known spatal features of the nterpolated phenomenon. These results are very promsng for an automated approach allowng on-demand datasets selecton and real-tme nterpolaton of prevously unknown measurements. Automaton s mportant n progressng towards a pan-european nterpolaton servce capablty. Keywords: Data Fuson, Krgng, Spatal Interpolaton, Wnd Speed, Wnd Drecton, OGC, WPS Introducton In-stu meteorologcal sensor measurements are generally recorded by sensor hardware at pont locatons, requrng some form of spatal nterpolaton should estmates at other locatons be needed. Many spatal nterpolaton methods exst, both determnstc and geo-statstcal, wth accuraces dependent on the nature of the observed phenomena, spatal densty of sensors, temporal frequency of samplng and the consstency and accuracy of measurement. In the SAY project [6] we have developed an nterpolaton algorthm for handlng wnd measurements. Our use case s to generate spatal grds of estmated wnd measurements for contnual nput nto a bathng water qualty model [3] and subsequent lve predcton of mcrobal contamnaton levels of bathng water at beaches. Mcrobal contamnaton s mportant nformaton for coast guards when makng the decson as to f, and when, to close publc beaches. Our algorthms are phenomena ndependent, and n SAY we have also successfully appled them to ar polluton and ground dsplacement measurements. We present a knowledge-asssted krgng algorthm appled to hstorcal wnd measurements from the UK meteorologcal offce (UKMO) dataset archves [9]. We report cross-valdated results for wnd speed mean error (ME), root mean squared error (RMSE) and the range of estmated values. These results are compared drectly to results for seven alternatve nterpolaton methods reported n [7] on the same dataset. Our knowledge-asssted krgng algorthm takes as nput knowledge about known areas of spatal coheson, ether dentfed by an expert or heurstcally calculated from the CGIAR-CSI GeoPortal SRTM (90m resoluton) Dgtal Elevaton Dataset [8]; mult-regon ordnary krgng s then used to compute varograms for each regon. The computaton of the varograms s asssted by metadata representng wnd phenomenon characterstcs. Ths algorthm s hosted wthn an Open Geospatal Consortum (OGC) sensor servce framework [5], showng how sensor processng servces can be setup to 'plug and process' dfferent sensor measurement datasets on-demand. We outlne n secton 2 relevant related work, descrbe our algorthm n secton 3 and expermental results n secton 4. We dscuss and conclude n sectons 5 and 6. 2 Related work In addton to ordnary krgng [], whch s outlned n the next secton, there are a number of varants such as unversal krgng [3], where a trend n data s assumed, and co-krgng [], where covarates are provded n the same area as the prmary sampled measurement to assst n predcton. Our novel automated varogram selecton and mult-regon ordnary krgng approach s another krgng varant, makng use of expert knowledge about the phenomenon characterstcs and the spatal cohesveness n known regons to mprove predcton accuracy. Many general spatal nterpolaton technques [2] exst, wth most relevant ones ncludng trend surface analyss, nverse dstance weghtng, local polynomal nterpolaton and thn plate splne. Untl [7] few works had benchmarked these technques thoroughly for wnd
2 measurement nterpolaton, and our work bulds on these results for drect comparson. 3 Knowledge-asssted ordnary krgng Krgng [] s a statstcal technque for nterpolaton of random phenomenon that uses a lnear combnaton of measured values at observed spatal locatons to estmate the value at an unobserved locaton of nterest. In contrast to other nterpolaton technques, lke nverse dstance weghtng and thn plane splne, t uses a model of the phenomenon s spatal correlaton encoded as a corrrelogram or a sem-varogram. Krgng s an exact nterpolator, so t respects the observed values at the observed locatons, whch dfferentates t from nterpolaton technques lke local polynomal nterpolaton and trend surface analyss. For the purposes of our work we have used ordnary krgng wth a semvarogram correlaton model. When usng a semvarogram, the spatal correlaton of the phenomenon s quantfed by means of a sem-varance functon (), where (h) s the number of pars of observed locatons s a dstance h apart, and z(s ) s the observed value at s. ( h) 2 ˆ γ ( h) = [ ( ) ( )] ( ) z s + h z s () 2 h = The sem-varogram can be a functon of both dstance and drecton, so t can account for drecton-dependent varablty (ansotropc spatal pattern). For krgng, a smooth sem-varogram s requred and we use smooth parametrc functons as approxmate models of the semvarogram [2]. These smooth functons are ftted to the sem-varance functon, or to the expermental semvarogram obtaned by averagng the sem-varance functon over a set of dstance lags, and after a goodness of ft analyss the best fttng functon s selected [3] [4]. We refer to ths functon as a 'varogram model' and the expermental sem-varogram as an 'expermental varogram'. In our automated varogram model selecton we use eght dfferent famles of varogram models: sphercal, exponental, Gaussan, lnear, power, generalsed Bessel, sne hole-effect and cosne hole-effect. It s usual that a phenomenon expert wll select a partcular varogram model based on experence and ft the selected model to the avalable observatons. Human expert nterventon n the nterpolaton process s expensve and makes the nterpolaton process phenomena specfc. Our soluton, ordnary krgng wth automated varogram model selecton (AVMS), tackles ths problem by utlsng metadata representng hgh level phenomenon characterstcs and automatng the varogram model selecton. 3. Automated nterpolaton workflow The varous offlne and real-tme stages of our knowledge-asssted nterpolaton process s shown n Fgure. Fgure. Ordnary krgng wth automated varogram model selecton procedure. The frst stage s the data pre-processng stage, where data cleansng, normalsaton and all necessary data transformatons are performed. The pre-processng stage ncludes nput of knowledge-based descrptons for regons of spatal coheson and offlne preparaton usng segmentaton technques and heurstcs pror to run-tme. In the data post-processng stage, data de-normalsaton and reverse transformatons are performed. The core stages of ordnary krgng wth AVMS are the expermental varogram creaton, varogram model selecton, model optmsaton and ordnary krgng. A varogram s created for each regon of spatal coheson. Phenomenon metadata s setup va expert defned profle confguratons. 3.2 Offlne dentfcaton of knowledge about spatally coherent regons Regon calculaton s performed automatcally from ether expert drawn Google Earth spatal polygons (KML format) or from ASA alttude data (ASC format) from the CGIAR-CSI GeoPortal SRTM (90m resoluton) Dgtal Elevaton Dataset [8]. The polygons / alttude maps are rendered as greyscale mages and standard mage processng technques (colour reducton, bnary mask per colour, pxel blur, labellng and edge dentfcaton) used to segment maps nto unque regons sutable for nput nto the krgng process. Regon segmentaton s executed offlne, va an automated web servce, as part of the ntal confguraton stage pror to on-demand krgng.
3 We expect over mountanous ground the mean daly wnd speed wll have lower levels of spatal correlaton than flat land. The land/sea boundary wll also have an effect, especally snce our sensors are land-based. Fgure 2 shows the alttude segmentaton, whch we found to be somewhat over-segmented and too fne-graned for our wnd phenomena. Fgure 3 shows the regon segmentaton for the expert drawn polygons, whch produced the best results and are used n the experments later n ths paper. Fgure 2. Regon segmentaton based on CGIAR-CSI GeoPortal SRTM alttude data Fgure 3. Regon segmentaton based on polygons drawn by an expert for land/water and flat/hlly areas The output of the offlne regon segmentaton process are a set of comma separated varable (CSV) maps contanng regon labels for every nterpolaton pont, and nterregon neghbour lnkage whch s used later by the knowledge-asssted ordnary krgng algorthm. 3.3 Real-tme automated varogram model creaton The most crtcal part of the expermental varogram creaton stage s the selecton of lags. Lags need to be selected so they contan an optmal number of ponts, do not smooth out physcal phenomenon characterstcs and avod nose. Generally the ntal slope of the varogram carres the most nformaton, so our frst few lags contan a smaller number of ponts. If no hole-effect s expected lags subsequent to the ntal slope may contan a large number of ponts. If a hole-effect s expected these lags shall contan a lower number of ponts so the hole-effect s not smoothed out. The relatve number of ponts n a lag s specfed n the metadata suppled to the nterpolaton. In the next stage, the varogram model selecton, we ft our eght varogram models to the expermental varogram. The model shape s governed by a subset of the followng parameters: nugget, correlaton range, power, hole and sll. We use a weghted least squares fttng method to select a model that best fts the expermental varogram. We ntroduce expert knowledge about phenomenon by constranng the ftted model types and parameters values and thus obtan a varogram model reflectng the characterstcs of the phenomenon of nterest. For daly mean wnd speed we expect low levels of nose to be present n the observatons because of the averagng used to calculate the means (see secton 4.) and a low level of rapd phenomenon fluctuatons. In the phenomenon metadata we thus set the nugget to be constraned to maxmum of 20% of the sll. We expect the spatal correlaton of the daly mean wnd speed to decrease very slowly wth ncreasng the dstance. In the phenomenon metadata we thus set the lower and upper bounds of the correlaton range parameter to be relatvely hgh, respectvely 25% and 75% of the maxmum dstance n the expermental varogram. We don t expect a holeeffect so do not select models wth a hole-effect. After selectng the varogram model, parameter optmsaton s performed by mnmsng the mean error (ME, see equaton (2)) and the root mean square error (RMSE, see equaton (3)) calculated over 0-folds of cross-valdaton. We use smplex optmsaton wth a loss functon of sum of ME and RMSE. The phenomenon constrants are reflected n the loss functon by addng a hgh penalty value when the model parameters are outsde the ranges specfed n the phenomenon metadata. ext, ordnary krgng wth a movng neghbourhood s performed, usng the optmsed varogram model, and krgng mean and error computed. The krgng mean s the estmated value of the phenomenon at the locaton of nterest. The krgng error s a measure of how good the observatons locaton confguraton s for estmaton at the locaton of nterest. Ordnary krgng wth a movng neghbourhood s able to pck out varatons of the phenomenon mean along the nterpolated area. As the nterpolated area s relatvely
4 large n respect to the phenomenon scale we expect varatons n the mean. Thus, n the phenomenon metadata we have specfed a constraned movng neghbourhood consstng of the closest observatons to be used; the value of neghbours was chosen after expermentaton comparng results wth dfferent values. Ths choce can be automated by ncludng an addtonal parameter n the model optmsaton stage, whch we ntend to mplement n future versons. A pont map showng regon labels for the nterpolated area s suppled to our ordnary krgng procedure so we know whch regon each nterpolated pont belongs to. The observatons belongng to a partcular regon are used for workng out the varogram of that regon. If there are not enough observatons avalable for a gven regon, observatons from the neghbourng regons are used n order to buld the raw varogram. The mnmum number observatons for the automated varogram estmaton s argmn(n*(n-)/2>3* lags ), where n s the number of observatons and lags s the number of lags n the varogram,.e. we have mnmum of 3 raw varogram ponts per lag. When computng estmates requrng observatons from a neghbourng regon we average the varograms of the home and the neghbourng regons. In the phenomenon metadata we have an nter-regon correlaton factor, r, rangng between 0 and, where 0 ndcates mnmal correlaton between the regons and ndcates hgh correlaton. We modulate averaged nterregon varograms by ncreasng the varogram values by a quantty equallng -r tmes the varogram sll, up to a value not larger than the sll. If the varogram doesn t have a sll, such as the power model, then varogram values are ncreased by a quantty equallng -r tmes the current varogram value. In ths way we ntroduce addtonal knowledge of factors nfluencng the phenomenon spatal correlaton. 4 Experments and results In the SAY project we have used our ordnary krgng wth AVMS for nterpolatng mcro-scale phenomena (e.g. ground dsplacement caused by underground tunnellng) and mn / meso-scale phenomena (e.g. wnd speed and drecton) yeldng meanngful results. In ths paper we drectly compare the performance of our algorthm aganst 7 well known algorthms and mplementatons on a benchmark data set. A defntve work n ths context s [7], where the performance of seven nterpolaton methods are compared on a dataset of daly mean wnd speeds obtaned from sensor data suppled by the UK meteorologcal offce (UKMO). Smlarly to [7] we have used leave-one-out cross valdaton and calculated the mean error (ME) and root mean square error (RMSE) as follows: RMSE = = ^ z ( s ) z( ) s where s the number of valdaton folds, z^(s ) s the estmated value and z(s ) s the observed value. 4. Data For our wnd speed experments we used the same dataset as [7] for 27 th of March 200, whch was orgnally obtaned from the UKMO. At ths date 89 sensors n England, Wales, South of Scotland and orthern Ireland reported at least 2 tmes, the readngs averaged to obtan daly mean wnd speeds. The daly mean wnd speeds vary from 2.2 to 3.6 m/s. Fgure 4 depcts the locatons and magntudes of the daly mean wnd speed observatons, where the sze of the dots s relatve to the magntude of the speed. 2 (3) Fgure 4. Locatons and relatve magntude of daly mean wnd speed observatons for 27 th of March 200. For our wnd drecton experments we used 27 th March 200 data from the UKMO dataset archve [9]. It contans 57 observatons wth daly mean wnd drectons. Fgure 5 depcts the locatons of the daly mean wnd drectons observatons, where the vectors show the drecton of the wnd. We apply our ordnary krgng wth AVMS to ths data and nterpolate on a grd wth a resoluton of 5km by 5km. For daly mean wnd speed we compare our results aganst the results publshed n [7]. In [7] no results are presented for daly mean wnd drectons (just wnd speed) so we just state our expermental results wthout comparson as a benchmark for other researchers. ME = = ^ z ( s ) z( ) s (2)
5 Fgure 5. Locatons of daly mean wnd drecton observatons for 27 th of March Daly mean wnd speed estmates benchmarkng In ths experment we compare our ordnary krgng wth AVMS algorthm aganst the 7 algorthms evaluated n [7]. We set our algorthm wth the followng parameters:. Expermental varogram lags percentles: 5, 0, 5, 20, 30, 40, 50, 60, 70, 80, 90, ugget boundares: 0 to 20% of the sll. 3. Correlaton range boundares: 25% to 75% of the maxmum varogram dstance. 4. Varogram models to use: sphercal, exponental, Gaussan, lnear, power or generalsed Bessel. Ordnary krgng - no regon knowledge used The frst experment s usng krgng wthout any knowledge of regons and nter-regon cohesveness. The daly mean wnd speed surface produced, shown n Fgure 6, s consstent wth the nput values over land, and s very smlar to the results found n [7]. Hgh wnd speeds are accurately estmated over the north, the north-east and north-west of England, and also over the Isle of Man. Low wnd speed are accurately estmated over the Mdlands, and the west and south-east of England. The krgng error, show n Fgure 7, s small where we have hgh concentraton of observatons and larger where there are smaller number of observatons. The RMSE, ME and the mn and max nterpolated values suggest that the performance of ordnary krgng wth AVMS s slghtly worse than the ordnary krgng n [7] and slghtly better than unversal krgng and local polynomal (see table ). It should be noted that our ordnary krgng wth AVMS needs only a mnmal nput and confguraton from users, unlke the classc manual approach used n [7]. Fgure 6. Daly mean wnd speed estmates [89 sensors, no regons, 5km x 5km grd resoluton, unt m/s] Fgure 7. Krgng error of wnd speed estmates [89 sensors, no regons, 5km x 5km grd resoluton]. otce how reported error s smoothed across coastlne and land. Ordnary krgng - coastlne regon knowledge provded Snce all sensors are land-based sea value estmates are n all probablty not very relable. Supplyng expert-drawn coastlne regon boundary knowledge can help n ths case. For our second experment we performed coastlne segmentaton on Google Earth drawn coastlne polygons and suppled the resultng regons to our ordnary krgng wth AVMS. For ths experment, the nter-regon correlaton factor r was set to 0. to ndcate that there are mnmal smlartes n the way sea and land measurements nfluence each other. The estmated mean wnd speed and the krgng error surfaces are shown n Fgure 8 and Fgure 9 respectvely. The n-land wnd speed estmates change only slghtly, wth the only notable dfferences n estmated land values over the Isle of Man. Comparng fgures 6 and 8 one can see that now there are more accurate hgh wnd speed values predcted as the manland observaton's nterference s dmnshed somewhat by the low nterregon correlaton factor. Ths result s also reflected n the algorthm's overall performance metrcs shown n table. The krgng error for the Isle of Man s very dfferent from the prevous krgng error wthout usng regon
6 knowledge; ths s clear comparng fgures 7 and 9. There s a clear jump n the krgng error on the land-sea borders ndcatng correctly that the wnd speed estmates over the sea areas are unrelable. regon. In contrast the spotty krgng error pattern n the northern regon, and n Wales and south-west of England, we nterpreted as needng much more dense observatons to acheve good estmates n these regons. Fgure 8. Daly mean wnd speed estmates [89 sensors, coastlne regons, 5km x 5km grd resoluton, unt m/s] Fgure 0. Daly mean wnd speed estmates [89 sensors, flat/hlly regons, 5km x 5km grd resoluton, unt m/s] Fgure 9. Krgng error of wnd speed estmates [89 sensors, coastlne regons, 5km x 5km grd resoluton]. otce the sharp error boundary at the coastlne. Ordnary krgng - flat/hlly regon knowledge provded For our last experment regardng wnd speed we provded doman knowledge about the dfferent terran topology (.e. flat land, hlly land, mountanous land), agan segmented nto regons. We used a hgher nter-regon correlaton factor of 0.9 snce these new regons are nland. Results wth these new regons are agan mproved and are comparable to ordnary krgng n [7] (see table ). The bas, ME, s decreased as s the RMSE and the mnmum and maxmum of the estmated values are also closer to those present n the raw data. The estmated wnd speed surface sown n Fgure 0 suggests that wth approprate regonalsaton the extremtes are more lkely to be dentfed but not smoothed out. The krgng error map, Fgure, now shows dstnct patterns n each of the dfferent regons. For example the krgng error over the Mdlands, South and South-East regon s very smooth, whch we nterpret as not needng very dense observatons to acheve good estmates n ths Fgure. Krgng error of wnd speed estmates [89 sensors, flat/hlly regons, 5km x 5km grd resoluton]. otce the sharp error boundares between spatal features and dfferent nter-regon error profles. 4.3 Daly mean wnd drecton estmaton Usng the same ordnary krgng wth AVMS setup as for the wnd speed experments we ran tests on the UKMO dataset for 27 th March 200 wth no regons defned. In our mplementaton of ordnary krgng we have a specal procedure [0] for handlng perodc values lke wnd drecton, whch ncludes vector rotaton and Cartesan transformaton and smulaton. Estmated wnd drecton vectors are vsualsed n Fgure 2. We use a 28km x 28km grd to make vsualsaton easer for arrow vectors. The ME and RMSE are gven n table 2. We fnd the ME, of degrees, and RMSE, of 3.66 degrees, n relatve terms very low consderng the worse case could be 80 degrees. The mean relatve absolute error (MRAE) s 6%. For our wnd speed experments MRAE vares from 22% to 23%.
7 For wnd speed we calculate MRAE as: MRAE = = ^ z ( s ) z( s ) / z( s ) and for wnd drecton we calculate MRAE as: MRAE ( perodc) = = ^ norm( z ( s ) z( s ) ) /80 where s the number of valdaton folds, z^(s ) s the estmated value, z(s ) s the observed value,. s the absolute value and norm() yelds a normalsed wnd drecton n the range of -80 to +80. The wnd drecton krgng error s shown n Fgure 3, wth normalsed values n the range of 0 to 360. Krgng error quantfes the observatons confguraton sutablty for estmatng at a partcular locaton and s not a measure of the estmaton error n tself. We are unaware of any other benchmark wnd drecton nterpolaton results so publsh here n the hope of provdng other researchers wth a benchmark result set. Method umber of Regons Inter-regon correlaton Exec-tme (mns) Krgng_ 0 n/a Krgng_2 4 coastlne 0. 8 Krgng_3 7 flat/hlly All methods use ordnary krgng wth AVMS Computer spec : 2 CPU's (2.4 GHz) wth 4 Gbyte RAM Method Mn estmate m/s Max estmate m/s RMS E m/s ME m/s Krgng_ Krgng_ Krgng_ Cokrgng [7] Ordnary krgng [7] Local polynomal [7] Unversal krgng [7] IDW [7] TPS [7] TSA [7] Table. Comparson of knowledge-asssted krgng to benchmark nterpolaton results reported n [7] Method ME RMSE degrees degrees Ordnary krgng wth AVMS Table 2. Wnd drecton estmate ME and RMSE [57 sensors, no regon, 28km x 28km grd resoluton] Fgure 2. Daly mean wnd drectons [57 sensors, no regon, 28km x 28km grd resoluton, unt degrees] Fgure 3. Krgng error for wnd drecton estmates [57 sensors, no regon, 28km x 28km grd resoluton] 5 Dscusson In our experments we have nvestgated methods to automatcally select varogram models and ways to ntroduce expert knowledge nto varogram calculaton process. Our results show the mpact usng dfferent levels of expert knowledge, about both the phenomenology and spatal areas of nterest, can have on estmaton error. Ths automated approach s n contrast to exstng methods [7] that rely on manual varogram tunng by experts. We thnk our approach offers a realstc route to provdng ondemand krgng servces that can nterpolate measurement datasets selected by users and that are unknown untl runtme. Ths level of automaton s becomng more mportant as we see European envronment agences sharng ncreasng amounts of sensor data under the ISPIRE drectve [5]. Such automaton, and a move away from manual tunng and confguraton, ncreases scalablty and could allow truly dynamc pan-european nterpolaton for wnd measurements and other phenomenology. The performance of our ordnary krgng wth AVMS algorthm s slghtly worse than that reported by ordnary
8 krgng n [7] wth expert tuned varogram. Partly ths s due to the expert's skll n manually tunng the varogram. However, our mplementaton s also only sotropc, whch accounts for some of the performance compared aganst the ansotropc ordnary krgng algorthm n [7]. We ntend to mplement ansotropc krgng n future releases. Where our knowledge-asssted approach offers the most mprovement, compared to basc ordnary krgng, s n the enhanced spatal defnton of the krgng error maps, and therefore mproved confdence n them. Ths s most clear n areas where no sensors are located (e.g. offshore n the sea). Ths result can be appled more wdely to other spatal features than hlls and coastlnes, at a varety of dfferent spatal scales, such as buldng footprnts and rver outlets. Though not the focus of our wnd nterpolaton paper, we have successfully nterpolated measured phenomenon ncludng ground dsplacement, water salnty and turbdty. 6 Concluson We have benchmarked a novel knowledge-asssted krgng algorthm that allows knowledge of regonal spatal coheson to be specfed and varograms calculated per regon. The varogram calculaton tself s automated and phenomenon specfc metadata allows us to confgure krgng for more than just wnd phenomenon. Spatal regons are created offlne automatcally by segmentaton of ether expert-drawn Google Earth polygons or ASA alttude data. Our use-case s to create wnd nterpolaton grds for nput nto a bathng water qualty model of mcrobal contamnaton, and subsequent decson support for beach attendants to make bathng rsk assessments. We benchmark our knowledge-asssted krgng algorthm aganst 7 other algorthms usng the same UK met-offce wnd measurement dataset reported n [7]. Wnd speed estmaton results are comparable, but not better than ordnary krgng, but the krgng error maps are much sharper and reflect the known spatal features better. We also provde results for wnd drecton nterpolaton. These results are very promsng consderng t s an automated approach allowng on-demand datasets to be selected and real-tme nterpolaton of measurements unknown a-pror. Automaton s mportant n moves towards a pan-european nterpolaton servce capablty makng use of European envronment agency data shared n complance wth the European ISPIRE drectve [5]. Ths work was funded by the European Commsson s IST Programme under contract FP6-IST SAY [6] References [] D.G. Krge, Two-dmensonal weghted average trend surfaces for ore-evaluaton, Journal of the South Afrcan Insttute of Mnng and Metallurgy 66: 3 38, 966 [2].A.C. Cresse, Fttng varogram models by weghted least squares, Journal of the Internatonal Assocaton for Mathematcal Geology 7: , 985 [3].A.C. Cresse, Statstcs for Spatal Data, John Wley and Sons: ew York, 993 [4] P.A. Burrough and R.A. McDonnell, Prncples of Geographcal Informaton Systems, Clarendon Press: Oxford, 998 [5] S.E. Mddleton (ed) et al, SAY Fuson and Modellng Archtecture, OGC Dscusson paper 0-00, 200 [6] SAY project, [7] W. Luo, M.C. Taylor and S.R. Parker, A comparson of spatal nterpolaton methods to estmate contnuous wnd speed surfaces usng rregularly dstrbuted data from England and Wales, Int. J. Clmatol, 28: , 2008 [8] CGIAR-CSI, GeoPortal SRTM (90m resoluton) Dgtal Elevaton Dataset, [9] UK Met Offce, MIDAS Land Surface Observaton Statons Data, [0] Z. Zlatev, S.E. Mddleton and G. Veres, Ordnary krgng for on-demand average wnd nterpolaton of nstu wnd sensor data, EWEC 2009, France, March 2009 [] A.G. Journel, C. Hujbregts, Mnng Geostatstcs, Academc Press: ew York, 978 [2].S. Lam, Spatal nterpolaton methods: a revew, The Amercan Cartographer 0: 29 35, 983 [3] Z.A. Sabeur (ed) et al, Development of envronmental nformaton tools for the predcton of water qualty rsks n bathng waters, Fnal Report to the Envronment Agency of England and Wales, Interreg ICREW project, Plot Acton 4: Forecastng bathng water qualty. Fnal Techncal Report/FIAL. 2st June 2006, pp28, 2006 [4] R.B. Stull, Meteorology for Scentsts and Engneers, 2nd Edton, Earth Scences seres, Cengage Learnng Publsher. ISB-3: , 999 [5] ISPIRE drectve, Drectve 2007/2/EC of the European Parlament, 4 March 2007
X- Chart Using ANOM Approach
ISSN 1684-8403 Journal of Statstcs Volume 17, 010, pp. 3-3 Abstract X- Chart Usng ANOM Approach Gullapall Chakravarth 1 and Chaluvad Venkateswara Rao Control lmts for ndvdual measurements (X) chart are
More informationContent Based Image Retrieval Using 2-D Discrete Wavelet with Texture Feature with Different Classifiers
IOSR Journal of Electroncs and Communcaton Engneerng (IOSR-JECE) e-issn: 78-834,p- ISSN: 78-8735.Volume 9, Issue, Ver. IV (Mar - Apr. 04), PP 0-07 Content Based Image Retreval Usng -D Dscrete Wavelet wth
More informationSimulation: Solving Dynamic Models ABE 5646 Week 11 Chapter 2, Spring 2010
Smulaton: Solvng Dynamc Models ABE 5646 Week Chapter 2, Sprng 200 Week Descrpton Readng Materal Mar 5- Mar 9 Evaluatng [Crop] Models Comparng a model wth data - Graphcal, errors - Measures of agreement
More informationImprovement of Spatial Resolution Using BlockMatching Based Motion Estimation and Frame. Integration
Improvement of Spatal Resoluton Usng BlockMatchng Based Moton Estmaton and Frame Integraton Danya Suga and Takayuk Hamamoto Graduate School of Engneerng, Tokyo Unversty of Scence, 6-3-1, Nuku, Katsuska-ku,
More informationA Binarization Algorithm specialized on Document Images and Photos
A Bnarzaton Algorthm specalzed on Document mages and Photos Ergna Kavalleratou Dept. of nformaton and Communcaton Systems Engneerng Unversty of the Aegean kavalleratou@aegean.gr Abstract n ths paper, a
More informationS1 Note. Basis functions.
S1 Note. Bass functons. Contents Types of bass functons...1 The Fourer bass...2 B-splne bass...3 Power and type I error rates wth dfferent numbers of bass functons...4 Table S1. Smulaton results of type
More informationPage 0 of 0 SPATIAL INTERPOLATION METHODS
Page 0 of 0 SPATIAL INTERPOLATION METHODS 2018 1. Introducton Spatal nterpolaton s the procedure to predct the value of attrbutes at unobserved ponts wthn a study regon usng exstng observatons (Waters,
More informationFeature Reduction and Selection
Feature Reducton and Selecton Dr. Shuang LIANG School of Software Engneerng TongJ Unversty Fall, 2012 Today s Topcs Introducton Problems of Dmensonalty Feature Reducton Statstc methods Prncpal Components
More informationProblem Definitions and Evaluation Criteria for Computational Expensive Optimization
Problem efntons and Evaluaton Crtera for Computatonal Expensve Optmzaton B. Lu 1, Q. Chen and Q. Zhang 3, J. J. Lang 4, P. N. Suganthan, B. Y. Qu 6 1 epartment of Computng, Glyndwr Unversty, UK Faclty
More informationSLAM Summer School 2006 Practical 2: SLAM using Monocular Vision
SLAM Summer School 2006 Practcal 2: SLAM usng Monocular Vson Javer Cvera, Unversty of Zaragoza Andrew J. Davson, Imperal College London J.M.M Montel, Unversty of Zaragoza. josemar@unzar.es, jcvera@unzar.es,
More informationTN348: Openlab Module - Colocalization
TN348: Openlab Module - Colocalzaton Topc The Colocalzaton module provdes the faclty to vsualze and quantfy colocalzaton between pars of mages. The Colocalzaton wndow contans a prevew of the two mages
More informationUser Authentication Based On Behavioral Mouse Dynamics Biometrics
User Authentcaton Based On Behavoral Mouse Dynamcs Bometrcs Chee-Hyung Yoon Danel Donghyun Km Department of Computer Scence Department of Computer Scence Stanford Unversty Stanford Unversty Stanford, CA
More informationCluster Analysis of Electrical Behavior
Journal of Computer and Communcatons, 205, 3, 88-93 Publshed Onlne May 205 n ScRes. http://www.scrp.org/ournal/cc http://dx.do.org/0.4236/cc.205.350 Cluster Analyss of Electrcal Behavor Ln Lu Ln Lu, School
More informationAPPLICATION OF A COMPUTATIONALLY EFFICIENT GEOSTATISTICAL APPROACH TO CHARACTERIZING VARIABLY SPACED WATER-TABLE DATA
RFr"W/FZD JAN 2 4 1995 OST control # 1385 John J Q U ~ M Argonne Natonal Laboratory Argonne, L 60439 Tel: 708-252-5357, Fax: 708-252-3 611 APPLCATON OF A COMPUTATONALLY EFFCENT GEOSTATSTCAL APPROACH TO
More informationSmoothing Spline ANOVA for variable screening
Smoothng Splne ANOVA for varable screenng a useful tool for metamodels tranng and mult-objectve optmzaton L. Rcco, E. Rgon, A. Turco Outlne RSM Introducton Possble couplng Test case MOO MOO wth Game Theory
More informationCS 534: Computer Vision Model Fitting
CS 534: Computer Vson Model Fttng Sprng 004 Ahmed Elgammal Dept of Computer Scence CS 534 Model Fttng - 1 Outlnes Model fttng s mportant Least-squares fttng Maxmum lkelhood estmaton MAP estmaton Robust
More informationA Semi-parametric Regression Model to Estimate Variability of NO 2
Envronment and Polluton; Vol. 2, No. 1; 2013 ISSN 1927-0909 E-ISSN 1927-0917 Publshed by Canadan Center of Scence and Educaton A Sem-parametrc Regresson Model to Estmate Varablty of NO 2 Meczysław Szyszkowcz
More informationAdjustment methods for differential measurement errors in multimode surveys
Adjustment methods for dfferental measurement errors n multmode surveys Salah Merad UK Offce for Natonal Statstcs ESSnet MM DCSS, Fnal Meetng Wesbaden, Germany, 4-5 September 2014 Outlne Introducton Stablsng
More informationTHE THEORY OF REGIONALIZED VARIABLES
CHAPTER 4 THE THEORY OF REGIONALIZED VARIABLES 4.1 Introducton It s ponted out by Armstrong (1998 : 16) that Matheron (1963b), realzng the sgnfcance of the spatal aspect of geostatstcal data, coned the
More informationWishing you all a Total Quality New Year!
Total Qualty Management and Sx Sgma Post Graduate Program 214-15 Sesson 4 Vnay Kumar Kalakband Assstant Professor Operatons & Systems Area 1 Wshng you all a Total Qualty New Year! Hope you acheve Sx sgma
More informationNAG Fortran Library Chapter Introduction. G10 Smoothing in Statistics
Introducton G10 NAG Fortran Lbrary Chapter Introducton G10 Smoothng n Statstcs Contents 1 Scope of the Chapter... 2 2 Background to the Problems... 2 2.1 Smoothng Methods... 2 2.2 Smoothng Splnes and Regresson
More informationHigh resolution 3D Tau-p transform by matching pursuit Weiping Cao* and Warren S. Ross, Shearwater GeoServices
Hgh resoluton 3D Tau-p transform by matchng pursut Wepng Cao* and Warren S. Ross, Shearwater GeoServces Summary The 3D Tau-p transform s of vtal sgnfcance for processng sesmc data acqured wth modern wde
More informationA MOVING MESH APPROACH FOR SIMULATION BUDGET ALLOCATION ON CONTINUOUS DOMAINS
Proceedngs of the Wnter Smulaton Conference M E Kuhl, N M Steger, F B Armstrong, and J A Jones, eds A MOVING MESH APPROACH FOR SIMULATION BUDGET ALLOCATION ON CONTINUOUS DOMAINS Mark W Brantley Chun-Hung
More informationLecture #15 Lecture Notes
Lecture #15 Lecture Notes The ocean water column s very much a 3-D spatal entt and we need to represent that structure n an economcal way to deal wth t n calculatons. We wll dscuss one way to do so, emprcal
More informationAn Entropy-Based Approach to Integrated Information Needs Assessment
Dstrbuton Statement A: Approved for publc release; dstrbuton s unlmted. An Entropy-Based Approach to ntegrated nformaton Needs Assessment June 8, 2004 Wllam J. Farrell Lockheed Martn Advanced Technology
More informationFEATURE EXTRACTION. Dr. K.Vijayarekha. Associate Dean School of Electrical and Electronics Engineering SASTRA University, Thanjavur
FEATURE EXTRACTION Dr. K.Vjayarekha Assocate Dean School of Electrcal and Electroncs Engneerng SASTRA Unversty, Thanjavur613 41 Jont Intatve of IITs and IISc Funded by MHRD Page 1 of 8 Table of Contents
More informationDetermining the Optimal Bandwidth Based on Multi-criterion Fusion
Proceedngs of 01 4th Internatonal Conference on Machne Learnng and Computng IPCSIT vol. 5 (01) (01) IACSIT Press, Sngapore Determnng the Optmal Bandwdth Based on Mult-crteron Fuson Ha-L Lang 1+, Xan-Mn
More informationClassifier Selection Based on Data Complexity Measures *
Classfer Selecton Based on Data Complexty Measures * Edth Hernández-Reyes, J.A. Carrasco-Ochoa, and J.Fco. Martínez-Trndad Natonal Insttute for Astrophyscs, Optcs and Electroncs, Lus Enrque Erro No.1 Sta.
More informationSimulation Based Analysis of FAST TCP using OMNET++
Smulaton Based Analyss of FAST TCP usng OMNET++ Umar ul Hassan 04030038@lums.edu.pk Md Term Report CS678 Topcs n Internet Research Sprng, 2006 Introducton Internet traffc s doublng roughly every 3 months
More informationCompiler Design. Spring Register Allocation. Sample Exercises and Solutions. Prof. Pedro C. Diniz
Compler Desgn Sprng 2014 Regster Allocaton Sample Exercses and Solutons Prof. Pedro C. Dnz USC / Informaton Scences Insttute 4676 Admralty Way, Sute 1001 Marna del Rey, Calforna 90292 pedro@s.edu Regster
More informationWavefront Reconstructor
A Dstrbuted Smplex B-Splne Based Wavefront Reconstructor Coen de Vsser and Mchel Verhaegen 14-12-201212 2012 Delft Unversty of Technology Contents Introducton Wavefront reconstructon usng Smplex B-Splnes
More informationSkew Angle Estimation and Correction of Hand Written, Textual and Large areas of Non-Textual Document Images: A Novel Approach
Angle Estmaton and Correcton of Hand Wrtten, Textual and Large areas of Non-Textual Document Images: A Novel Approach D.R.Ramesh Babu Pyush M Kumat Mahesh D Dhannawat PES Insttute of Technology Research
More information5.0 Quality Assurance
5.0 Dr. Fred Omega Garces Analytcal Chemstry 25 Natural Scence, Mramar College Bascs of s what we do to get the rght answer for our purpose QA s planned and refers to planned and systematc producton processes
More informationA Fast Content-Based Multimedia Retrieval Technique Using Compressed Data
A Fast Content-Based Multmeda Retreval Technque Usng Compressed Data Borko Furht and Pornvt Saksobhavvat NSF Multmeda Laboratory Florda Atlantc Unversty, Boca Raton, Florda 3343 ABSTRACT In ths paper,
More informationParallelism for Nested Loops with Non-uniform and Flow Dependences
Parallelsm for Nested Loops wth Non-unform and Flow Dependences Sam-Jn Jeong Dept. of Informaton & Communcaton Engneerng, Cheonan Unversty, 5, Anseo-dong, Cheonan, Chungnam, 330-80, Korea. seong@cheonan.ac.kr
More informationSupport Vector Machines
/9/207 MIST.6060 Busness Intellgence and Data Mnng What are Support Vector Machnes? Support Vector Machnes Support Vector Machnes (SVMs) are supervsed learnng technques that analyze data and recognze patterns.
More informationThe Research of Support Vector Machine in Agricultural Data Classification
The Research of Support Vector Machne n Agrcultural Data Classfcaton Le Sh, Qguo Duan, Xnmng Ma, Me Weng College of Informaton and Management Scence, HeNan Agrcultural Unversty, Zhengzhou 45000 Chna Zhengzhou
More informationSynthesizer 1.0. User s Guide. A Varying Coefficient Meta. nalytic Tool. Z. Krizan Employing Microsoft Excel 2007
Syntheszer 1.0 A Varyng Coeffcent Meta Meta-Analytc nalytc Tool Employng Mcrosoft Excel 007.38.17.5 User s Gude Z. Krzan 009 Table of Contents 1. Introducton and Acknowledgments 3. Operatonal Functons
More informationImage Representation & Visualization Basic Imaging Algorithms Shape Representation and Analysis. outline
mage Vsualzaton mage Vsualzaton mage Representaton & Vsualzaton Basc magng Algorthms Shape Representaton and Analyss outlne mage Representaton & Vsualzaton Basc magng Algorthms Shape Representaton and
More informationA New Approach For the Ranking of Fuzzy Sets With Different Heights
New pproach For the ankng of Fuzzy Sets Wth Dfferent Heghts Pushpnder Sngh School of Mathematcs Computer pplcatons Thapar Unversty, Patala-7 00 Inda pushpndersnl@gmalcom STCT ankng of fuzzy sets plays
More informationAn Image Fusion Approach Based on Segmentation Region
Rong Wang, L-Qun Gao, Shu Yang, Yu-Hua Cha, and Yan-Chun Lu An Image Fuson Approach Based On Segmentaton Regon An Image Fuson Approach Based on Segmentaton Regon Rong Wang, L-Qun Gao, Shu Yang 3, Yu-Hua
More informationSubspace clustering. Clustering. Fundamental to all clustering techniques is the choice of distance measure between data points;
Subspace clusterng Clusterng Fundamental to all clusterng technques s the choce of dstance measure between data ponts; D q ( ) ( ) 2 x x = x x, j k = 1 k jk Squared Eucldean dstance Assumpton: All features
More informationMachine Learning 9. week
Machne Learnng 9. week Mappng Concept Radal Bass Functons (RBF) RBF Networks 1 Mappng It s probably the best scenaro for the classfcaton of two dataset s to separate them lnearly. As you see n the below
More informationNUMERICAL SOLVING OPTIMAL CONTROL PROBLEMS BY THE METHOD OF VARIATIONS
ARPN Journal of Engneerng and Appled Scences 006-017 Asan Research Publshng Network (ARPN). All rghts reserved. NUMERICAL SOLVING OPTIMAL CONTROL PROBLEMS BY THE METHOD OF VARIATIONS Igor Grgoryev, Svetlana
More informationOptimizing Document Scoring for Query Retrieval
Optmzng Document Scorng for Query Retreval Brent Ellwen baellwe@cs.stanford.edu Abstract The goal of ths project was to automate the process of tunng a document query engne. Specfcally, I used machne learnng
More informationType-2 Fuzzy Non-uniform Rational B-spline Model with Type-2 Fuzzy Data
Malaysan Journal of Mathematcal Scences 11(S) Aprl : 35 46 (2017) Specal Issue: The 2nd Internatonal Conference and Workshop on Mathematcal Analyss (ICWOMA 2016) MALAYSIAN JOURNAL OF MATHEMATICAL SCIENCES
More informationTerm Weighting Classification System Using the Chi-square Statistic for the Classification Subtask at NTCIR-6 Patent Retrieval Task
Proceedngs of NTCIR-6 Workshop Meetng, May 15-18, 2007, Tokyo, Japan Term Weghtng Classfcaton System Usng the Ch-square Statstc for the Classfcaton Subtask at NTCIR-6 Patent Retreval Task Kotaro Hashmoto
More informationA mathematical programming approach to the analysis, design and scheduling of offshore oilfields
17 th European Symposum on Computer Aded Process Engneerng ESCAPE17 V. Plesu and P.S. Agach (Edtors) 2007 Elsever B.V. All rghts reserved. 1 A mathematcal programmng approach to the analyss, desgn and
More informationPositive Semi-definite Programming Localization in Wireless Sensor Networks
Postve Sem-defnte Programmng Localzaton n Wreless Sensor etworks Shengdong Xe 1,, Jn Wang, Aqun Hu 1, Yunl Gu, Jang Xu, 1 School of Informaton Scence and Engneerng, Southeast Unversty, 10096, anjng Computer
More informationSteps for Computing the Dissimilarity, Entropy, Herfindahl-Hirschman and. Accessibility (Gravity with Competition) Indices
Steps for Computng the Dssmlarty, Entropy, Herfndahl-Hrschman and Accessblty (Gravty wth Competton) Indces I. Dssmlarty Index Measurement: The followng formula can be used to measure the evenness between
More informationCourse Introduction. Algorithm 8/31/2017. COSC 320 Advanced Data Structures and Algorithms. COSC 320 Advanced Data Structures and Algorithms
Course Introducton Course Topcs Exams, abs, Proects A quc loo at a few algorthms 1 Advanced Data Structures and Algorthms Descrpton: We are gong to dscuss algorthm complexty analyss, algorthm desgn technques
More informationSupport Vector Machines
Support Vector Machnes Decson surface s a hyperplane (lne n 2D) n feature space (smlar to the Perceptron) Arguably, the most mportant recent dscovery n machne learnng In a nutshell: map the data to a predetermned
More informationResearch and Application of Fingerprint Recognition Based on MATLAB
Send Orders for Reprnts to reprnts@benthamscence.ae The Open Automaton and Control Systems Journal, 205, 7, 07-07 Open Access Research and Applcaton of Fngerprnt Recognton Based on MATLAB Nng Lu* Department
More informationLearning the Kernel Parameters in Kernel Minimum Distance Classifier
Learnng the Kernel Parameters n Kernel Mnmum Dstance Classfer Daoqang Zhang 1,, Songcan Chen and Zh-Hua Zhou 1* 1 Natonal Laboratory for Novel Software Technology Nanjng Unversty, Nanjng 193, Chna Department
More informationCell Count Method on a Network with SANET
CSIS Dscusson Paper No.59 Cell Count Method on a Network wth SANET Atsuyuk Okabe* and Shno Shode** Center for Spatal Informaton Scence, Unversty of Tokyo 7-3-1, Hongo, Bunkyo-ku, Tokyo 113-8656, Japan
More informationA Robust Method for Estimating the Fundamental Matrix
Proc. VIIth Dgtal Image Computng: Technques and Applcatons, Sun C., Talbot H., Ourseln S. and Adraansen T. (Eds.), 0- Dec. 003, Sydney A Robust Method for Estmatng the Fundamental Matrx C.L. Feng and Y.S.
More informationRadial Basis Functions
Radal Bass Functons Mesh Reconstructon Input: pont cloud Output: water-tght manfold mesh Explct Connectvty estmaton Implct Sgned dstance functon estmaton Image from: Reconstructon and Representaton of
More information2x x l. Module 3: Element Properties Lecture 4: Lagrange and Serendipity Elements
Module 3: Element Propertes Lecture : Lagrange and Serendpty Elements 5 In last lecture note, the nterpolaton functons are derved on the bass of assumed polynomal from Pascal s trangle for the fled varable.
More informationAvailable online at ScienceDirect. Procedia Environmental Sciences 26 (2015 )
Avalable onlne at www.scencedrect.com ScenceDrect Proceda Envronmental Scences 26 (2015 ) 109 114 Spatal Statstcs 2015: Emergng Patterns Calbratng a Geographcally Weghted Regresson Model wth Parameter-Specfc
More informationHelsinki University Of Technology, Systems Analysis Laboratory Mat Independent research projects in applied mathematics (3 cr)
Helsnk Unversty Of Technology, Systems Analyss Laboratory Mat-2.08 Independent research projects n appled mathematcs (3 cr) "! #$&% Antt Laukkanen 506 R ajlaukka@cc.hut.f 2 Introducton...3 2 Multattrbute
More informationLearning-Based Top-N Selection Query Evaluation over Relational Databases
Learnng-Based Top-N Selecton Query Evaluaton over Relatonal Databases Lang Zhu *, Wey Meng ** * School of Mathematcs and Computer Scence, Hebe Unversty, Baodng, Hebe 071002, Chna, zhu@mal.hbu.edu.cn **
More informationLobachevsky State University of Nizhni Novgorod. Polyhedron. Quick Start Guide
Lobachevsky State Unversty of Nzhn Novgorod Polyhedron Quck Start Gude Nzhn Novgorod 2016 Contents Specfcaton of Polyhedron software... 3 Theoretcal background... 4 1. Interface of Polyhedron... 6 1.1.
More informationA Fast Visual Tracking Algorithm Based on Circle Pixels Matching
A Fast Vsual Trackng Algorthm Based on Crcle Pxels Matchng Zhqang Hou hou_zhq@sohu.com Chongzhao Han czhan@mal.xjtu.edu.cn Ln Zheng Abstract: A fast vsual trackng algorthm based on crcle pxels matchng
More informationEdge Detection in Noisy Images Using the Support Vector Machines
Edge Detecton n Nosy Images Usng the Support Vector Machnes Hlaro Gómez-Moreno, Saturnno Maldonado-Bascón, Francsco López-Ferreras Sgnal Theory and Communcatons Department. Unversty of Alcalá Crta. Madrd-Barcelona
More informationMeta-heuristics for Multidimensional Knapsack Problems
2012 4th Internatonal Conference on Computer Research and Development IPCSIT vol.39 (2012) (2012) IACSIT Press, Sngapore Meta-heurstcs for Multdmensonal Knapsack Problems Zhbao Man + Computer Scence Department,
More informationDetection of an Object by using Principal Component Analysis
Detecton of an Object by usng Prncpal Component Analyss 1. G. Nagaven, 2. Dr. T. Sreenvasulu Reddy 1. M.Tech, Department of EEE, SVUCE, Trupath, Inda. 2. Assoc. Professor, Department of ECE, SVUCE, Trupath,
More informationActive Contours/Snakes
Actve Contours/Snakes Erkut Erdem Acknowledgement: The sldes are adapted from the sldes prepared by K. Grauman of Unversty of Texas at Austn Fttng: Edges vs. boundares Edges useful sgnal to ndcate occludng
More informationProgramming in Fortran 90 : 2017/2018
Programmng n Fortran 90 : 2017/2018 Programmng n Fortran 90 : 2017/2018 Exercse 1 : Evaluaton of functon dependng on nput Wrte a program who evaluate the functon f (x,y) for any two user specfed values
More informationComputer Animation and Visualisation. Lecture 4. Rigging / Skinning
Computer Anmaton and Vsualsaton Lecture 4. Rggng / Sknnng Taku Komura Overvew Sknnng / Rggng Background knowledge Lnear Blendng How to decde weghts? Example-based Method Anatomcal models Sknnng Assume
More informationReview of approximation techniques
CHAPTER 2 Revew of appromaton technques 2. Introducton Optmzaton problems n engneerng desgn are characterzed by the followng assocated features: the objectve functon and constrants are mplct functons evaluated
More informationDesign of Georeference-Based Emission Activity Modeling System (G-BEAMS) for Japanese Emission Inventory Management
1 13 th Internatonal Emsson Inventory Conference June 7-10, 2004 Clearwater, Florda Sesson 7 Data Management Desgn of Georeference-Based Emsson Actvty Modelng System (G-BEAMS) for Japanese Emsson Inventory
More informationRelevance Assignment and Fusion of Multiple Learning Methods Applied to Remote Sensing Image Analysis
Assgnment and Fuson of Multple Learnng Methods Appled to Remote Sensng Image Analyss Peter Bajcsy, We-Wen Feng and Praveen Kumar Natonal Center for Supercomputng Applcaton (NCSA), Unversty of Illnos at
More informationAccuracy Assessment and Comparative Analysis of IDW, Spline and Kriging in Spatial Interpolation of Landform (Topography): An Experimental Study
Journal of Geographc Informaton System, 207, 9, 354-37 http://www.scrp.org/journal/jgs ISSN Onlne: 25-969 ISSN Prnt: 25-950 Accuracy Assessment and Comparatve Analyss of IDW, Splne and Krgng n Spatal Interpolaton
More informationy and the total sum of
Lnear regresson Testng for non-lnearty In analytcal chemstry, lnear regresson s commonly used n the constructon of calbraton functons requred for analytcal technques such as gas chromatography, atomc absorpton
More informationFuzzy Logic Based RS Image Classification Using Maximum Likelihood and Mahalanobis Distance Classifiers
Research Artcle Internatonal Journal of Current Engneerng and Technology ISSN 77-46 3 INPRESSCO. All Rghts Reserved. Avalable at http://npressco.com/category/jcet Fuzzy Logc Based RS Image Usng Maxmum
More informationPerformance Evaluation of Information Retrieval Systems
Why System Evaluaton? Performance Evaluaton of Informaton Retreval Systems Many sldes n ths secton are adapted from Prof. Joydeep Ghosh (UT ECE) who n turn adapted them from Prof. Dk Lee (Unv. of Scence
More informationAn efficient method to build panoramic image mosaics
An effcent method to buld panoramc mage mosacs Pattern Recognton Letters vol. 4 003 Dae-Hyun Km Yong-In Yoon Jong-Soo Cho School of Electrcal Engneerng and Computer Scence Kyungpook Natonal Unv. Abstract
More informationThe Codesign Challenge
ECE 4530 Codesgn Challenge Fall 2007 Hardware/Software Codesgn The Codesgn Challenge Objectves In the codesgn challenge, your task s to accelerate a gven software reference mplementaton as fast as possble.
More informationData Mining: Model Evaluation
Data Mnng: Model Evaluaton Aprl 16, 2013 1 Issues: Evaluatng Classfcaton Methods Accurac classfer accurac: predctng class label predctor accurac: guessng value of predcted attrbutes Speed tme to construct
More informationSolutions to Programming Assignment Five Interpolation and Numerical Differentiation
College of Engneerng and Coputer Scence Mechancal Engneerng Departent Mechancal Engneerng 309 Nuercal Analyss of Engneerng Systes Sprng 04 Nuber: 537 Instructor: Larry Caretto Solutons to Prograng Assgnent
More informationFitting: Deformable contours April 26 th, 2018
4/6/08 Fttng: Deformable contours Aprl 6 th, 08 Yong Jae Lee UC Davs Recap so far: Groupng and Fttng Goal: move from array of pxel values (or flter outputs) to a collecton of regons, objects, and shapes.
More informationEECS 730 Introduction to Bioinformatics Sequence Alignment. Luke Huan Electrical Engineering and Computer Science
EECS 730 Introducton to Bonformatcs Sequence Algnment Luke Huan Electrcal Engneerng and Computer Scence http://people.eecs.ku.edu/~huan/ HMM Π s a set of states Transton Probabltes a kl Pr( l 1 k Probablty
More informationModeling Local Uncertainty accounting for Uncertainty in the Data
Modelng Local Uncertanty accontng for Uncertanty n the Data Olena Babak and Clayton V Detsch Consder the problem of estmaton at an nsampled locaton sng srrondng samples The standard approach to ths problem
More informationQuery Clustering Using a Hybrid Query Similarity Measure
Query clusterng usng a hybrd query smlarty measure Fu. L., Goh, D.H., & Foo, S. (2004). WSEAS Transacton on Computers, 3(3), 700-705. Query Clusterng Usng a Hybrd Query Smlarty Measure Ln Fu, Don Hoe-Lan
More informationAssignment # 2. Farrukh Jabeen Algorithms 510 Assignment #2 Due Date: June 15, 2009.
Farrukh Jabeen Algorthms 51 Assgnment #2 Due Date: June 15, 29. Assgnment # 2 Chapter 3 Dscrete Fourer Transforms Implement the FFT for the DFT. Descrbed n sectons 3.1 and 3.2. Delverables: 1. Concse descrpton
More informationR s s f. m y s. SPH3UW Unit 7.3 Spherical Concave Mirrors Page 1 of 12. Notes
SPH3UW Unt 7.3 Sphercal Concave Mrrors Page 1 of 1 Notes Physcs Tool box Concave Mrror If the reflectng surface takes place on the nner surface of the sphercal shape so that the centre of the mrror bulges
More informationLECTURE : MANIFOLD LEARNING
LECTURE : MANIFOLD LEARNING Rta Osadchy Some sldes are due to L.Saul, V. C. Raykar, N. Verma Topcs PCA MDS IsoMap LLE EgenMaps Done! Dmensonalty Reducton Data representaton Inputs are real-valued vectors
More informationPRÉSENTATIONS DE PROJETS
PRÉSENTATIONS DE PROJETS Rex Onlne (V. Atanasu) What s Rex? Rex s an onlne browser for collectons of wrtten documents [1]. Asde ths core functon t has however many other applcatons that make t nterestng
More informationTsinghua University at TAC 2009: Summarizing Multi-documents by Information Distance
Tsnghua Unversty at TAC 2009: Summarzng Mult-documents by Informaton Dstance Chong Long, Mnle Huang, Xaoyan Zhu State Key Laboratory of Intellgent Technology and Systems, Tsnghua Natonal Laboratory for
More informationA Unified Framework for Semantics and Feature Based Relevance Feedback in Image Retrieval Systems
A Unfed Framework for Semantcs and Feature Based Relevance Feedback n Image Retreval Systems Ye Lu *, Chunhu Hu 2, Xngquan Zhu 3*, HongJang Zhang 2, Qang Yang * School of Computng Scence Smon Fraser Unversty
More informationMULTISPECTRAL IMAGES CLASSIFICATION BASED ON KLT AND ATR AUTOMATIC TARGET RECOGNITION
MULTISPECTRAL IMAGES CLASSIFICATION BASED ON KLT AND ATR AUTOMATIC TARGET RECOGNITION Paulo Quntlano 1 & Antono Santa-Rosa 1 Federal Polce Department, Brasla, Brazl. E-mals: quntlano.pqs@dpf.gov.br and
More informationS.P.H. : A SOLUTION TO AVOID USING EROSION CRITERION?
S.P.H. : A SOLUTION TO AVOID USING EROSION CRITERION? Célne GALLET ENSICA 1 place Emle Bloun 31056 TOULOUSE CEDEX e-mal :cgallet@ensca.fr Jean Luc LACOME DYNALIS Immeuble AEROPOLE - Bat 1 5, Avenue Albert
More informationIP Camera Configuration Software Instruction Manual
IP Camera 9483 - Confguraton Software Instructon Manual VBD 612-4 (10.14) Dear Customer, Wth your purchase of ths IP Camera, you have chosen a qualty product manufactured by RADEMACHER. Thank you for the
More informationA Statistical Model Selection Strategy Applied to Neural Networks
A Statstcal Model Selecton Strategy Appled to Neural Networks Joaquín Pzarro Elsa Guerrero Pedro L. Galndo joaqun.pzarro@uca.es elsa.guerrero@uca.es pedro.galndo@uca.es Dpto Lenguajes y Sstemas Informátcos
More informationDesign of a Real Time FPGA-based Three Dimensional Positioning Algorithm
Desgn of a Real Tme FPGA-based Three Dmensonal Postonng Algorthm Nathan G. Johnson-Wllams, Student Member IEEE, Robert S. Myaoka, Member IEEE, Xaol L, Student Member IEEE, Tom K. Lewellen, Fellow IEEE,
More informationLife Tables (Times) Summary. Sample StatFolio: lifetable times.sgp
Lfe Tables (Tmes) Summary... 1 Data Input... 2 Analyss Summary... 3 Survval Functon... 5 Log Survval Functon... 6 Cumulatve Hazard Functon... 7 Percentles... 7 Group Comparsons... 8 Summary The Lfe Tables
More informationImplementation Naïve Bayes Algorithm for Student Classification Based on Graduation Status
Internatonal Journal of Appled Busness and Informaton Systems ISSN: 2597-8993 Vol 1, No 2, September 2017, pp. 6-12 6 Implementaton Naïve Bayes Algorthm for Student Classfcaton Based on Graduaton Status
More informationAn Optimal Algorithm for Prufer Codes *
J. Software Engneerng & Applcatons, 2009, 2: 111-115 do:10.4236/jsea.2009.22016 Publshed Onlne July 2009 (www.scrp.org/journal/jsea) An Optmal Algorthm for Prufer Codes * Xaodong Wang 1, 2, Le Wang 3,
More informationEmpirical Distributions of Parameter Estimates. in Binary Logistic Regression Using Bootstrap
Int. Journal of Math. Analyss, Vol. 8, 4, no. 5, 7-7 HIKARI Ltd, www.m-hkar.com http://dx.do.org/.988/jma.4.494 Emprcal Dstrbutons of Parameter Estmates n Bnary Logstc Regresson Usng Bootstrap Anwar Ftranto*
More informationOutline. Type of Machine Learning. Examples of Application. Unsupervised Learning
Outlne Artfcal Intellgence and ts applcatons Lecture 8 Unsupervsed Learnng Professor Danel Yeung danyeung@eee.org Dr. Patrck Chan patrckchan@eee.org South Chna Unversty of Technology, Chna Introducton
More information