住宅本身特性和鄰里環境為影響住宅價格價格高低之主要因素，環境特徵通因素常透過公共設施反映，設施越完善表示鄰里生活機能及可及性越高，產生正面效應使得住宅價格提升；鄰避設施產生的負面外部性導致居住品質下降，造成住宅價格降低。然現有文獻多數偏向針對特定的公共設施類型分析，相對少見到同時針對多樣設施進行全面性研究，缺乏生活設施面向的實證及設施服務水準衡量。 本研究運用特徵價格理論結合地理資訊系統，界定多項重大設施與生活設施；以等級衡量生活設施服務水準，並建構最小平方及分量迴歸模型自動分群探討設施之量增趨勢對住宅總價影響，以對設施作全面性研究。本研究採用臺北市2011年7月至2015年9月不動產交易實價登錄樣本與設施資料為對象。實證結果顯示，住宅總價主要受交通便利性和開放空間效益影響，生活設施水準對總價為正向影響，顯示生活機能具有正面效益。設施影響係數依序為捷運站0.0774、主題公園0.0307、餐飲0.0283、大專院校0.0219、大型醫院0.0193、殯葬設施-0.0190、學校0.0124、市場0.0057、鄰里公園0.0042及便利商店-0.0018。本研究進而探討不同總價下的設施影響效果發現，隨著住宅總價從低到高，各項生活設施的影響下降，低總價住宅主要受到生活機能和開放空間等鄰里環境影響，設施的便利性大於可及性效益；高總價住宅以負面外部性、文教及開放空間效益和就醫便利為主要影響，設施的便利效益不顯著。 / House pricing is strongly affected by its characteristics and neighborhood environment. Neighborhood characteristics are usually reflected by public facilities. The better the facilities are, the higher quality and convenience of the life results in a positive effect on housing price. The negative externalities of the NIMBY (not-in-my-back-yard) facilities lead to a decline in the quality of living and a reduction in housing prices. However, most of the existing literatures tend to focus on specific types of public facility. In this paper, more detailed study on the effects of overall facilities is proposed. In this research, the overall influence of facilities service level on house pricing is the focus. The proposed regression models are based on Hedonic price theory combined with geographic information system (GIS) to automatically cluster (or quantize) the facility numbers. The clustering is used for evaluating the relationships between facility number and housing price. Those overall facilities are categorized into two groups, important and community facilities. The provided different service levels of community facilities are defined. The real estate sales database in Taipei City from July 2012 to Sept. 2015 along with facilities is used for the evaluations. The evaluated results show that the total price is mainly affected by the convenience of transport and open space benefits, the service levels of community facilities have a positive effect. The impacts of the facilities are in the order of MRT, theme parks, restaurants, Universities, large hospitals, funeral facilities, schools, markets, neighborhood parks and convenience stores. Theme parks, the service level of community facilities and universities have higher impacts on those low-priced property while. funeral facilities, theme parks and large hospitals have higher impacts on high-priced one. One more finding is that the impacts of the service level of community facilities has declined with the increase in total housing prices. Low-priced property is mainly affected by the neighborhood environment and open space because facilities convenience is greater than the accessibility. High-priced property is mainly affected by the negative externalities, education, open space and access to medical care.

現行公告土地現值的評估，係採用人工的傳統方法，估價結果誤差甚大且過於主觀，無法達到大量估價客觀、快速、精確之目標。本文首先利用土地經濟理論的分析，探討土地市場價值、交易價格、評估價值之間的關係，釐清常見的混淆概念。並藉由估價比率研究，討論公告現值與市價差距的檢定模型，針對台南市的實際資料進行統計檢定，結果發現平均估價比率落於46.74﹪~48.52﹪之區間，並存在輕微的垂直不公平。 為改進現行公告現值不夠準確之缺失，本研究基於都市經濟理論與估價先驗訊息之基礎，利用特徵價格法與可加性模型建立大量估價模式。實證結果發現，影響台南市地價之因素，以區位、臨街關係、路寬、使用分區最為重要。在部份年度中，亦證實存在基地面積規模不經濟（plattage）現象。 傳統特徵價格法必須預設函數型態，若函數設定錯誤則將使參數估計產生偏誤。可加性模型結合無母數迴歸與母數迴歸之優點，不須預設函數型態、估計結果易於解釋且維持母數迴歸之收斂速度。其可經由修勻法配適出更客觀的函數關係，無論在樣本內與樣本外之估計均較特徵價格法為佳。 研究結果發現，本文所提出的二種估價模式確可達到快速精確的目標，使估價比率接近１，比目前評估效率提高一倍；在公平性方面雖無改善，但亦無嚴重之垂直不公平。其中可加性模型又較特徵價格法為佳，在電腦技術快速進步的今天，應用至大量估價的可行性大為提高，值得後續進一步深入研究。 / The present Announced Land Current Value （ALCV）was evaluated by traditional appraisal method that may result in large errors. Comparing to mass assessment approaches, it is hard to be objective, quick and precise. This research begins with the analysis based on land economic theory to discuss the relation among the market value, sale price and assessed value of land in order to clarify some confusing concepts. Through assessment-sale price ratio study, we analyze the difference between ALCV and sale price, and then use the actual data of Tainan City for empirical study. The results show that the average a-s ratio falls between 46.74%~48.52% with slight vertical inequity. To improve the lack of preciseness and objectivity of the present ALCV, this research uses hedonic price theory and Generalized Additive Model（GAM）based on urban economic theory and appraisal priori information. The results show that location, relations with adjacent streets, road width and zoning are the most influencing factors of land price in Tainan City. During some years, the phenomenon of plattage effect also exits. The function form must be set beforehand in the traditional hedonic pricing, meanwhile parameters bias will occur if the pre-determined function form were wrong. GAM has the advantages of nonparametric regression and parametric regression. The function form needs not to be pre-determined, the empirical results are easy to interpret, and the speed of variable convergence can be maintained. More precise functional relations can also be smoothed by GAM. It is superior to the traditional hedonic price in the sample and out of the sample prediction alike. The results of empirical study show that both of two models can reach the goal of rapidity and preciseness and make the a-s ratio toward 1. As to the equity, although they are not improved very much, the models don't bring serious vertical inequity. However, GAM is better than hedonic pricing when compared to each other. Due to the great progress of computer technology, the application of GAM to mass assessment can be increased greatly and is worthy continuing further study.

過去由於國內住宅市場的資訊並不流通，以致市場上出現的價格資訊相當混亂，就價格的種類來看，各種名目住宅價格間的差異未有明確釐清，使一般人常會對不同的住宅價格產生誤解。就時間序列來看，不同的時間、地區及住宅類型，到底住宅價格的變動如何，也未能有一嚴謹且量化之指標表示。 　　本研究運用標準住宅的概念編製住宅價格指數，所謂標準住宅乃是指一定時間、地區、類型，市場上成交的住宅中，典型的住宅屬性及其數量的組合，也就是說，觀察市場上成交典型住宅的價格變動情形，作為指數編製的基礎，以控制住宅的異質性，再以特徵價格法來求得各屬性的單價以進一步控制品質。另外，由於住宅成交數量變動較大，因此運用裴氏公式作為指數公式，使加權權重的誤差不致太大。 　　由各地區的指數變動趨勢可得到以下幾點結論：首先，一般所認為的三次房地產價格高峰期（62至63年、68至70年、76至78年），從指數的變動來看並不明顯，只有76至78年的上漲趨勢較明顯，在經過幣值平減後，長期趨勢更顯平緩；其次，長期來看，住宅價格持續上漲，部分時期持平或下跌，但幅度及持續時間有限，故所謂房價下跌，其實跌的是上漲率；最後，就上漲幅度來看，台北縣、市的幅度最大，除台北市外，非都市地區（非省轄市）房價的上漲速度較都市地區（省轄市）為高。 　　最後歸納二個造成一般人對住宅價格變動之錯誤印象的原因。主要是品質未加控制，由於品質會影響住宅價格，且消費的住宅品質及數量會隨時間而改變，一般人未察覺此點，而造成對房價上漲的誤解。其次是未考慮幣值，由於「今天的一塊錢不等於明天的一塊錢」，因此以名目價格觀察住宅價格變動的作法，也會造成對房價變動的誤解。

市地重劃係依照都市發展趨勢及都市計畫規劃內容，將一定範圍內畸零細碎不規則之土地，依據法令加以重新整理，重劃區內土地除配合公共設施興建後，其餘可建築用地依原位次分配原則、調整分配原則並留設抵費地後，重新分配予原土地所有權人，其最終目的是地權的調整，而地權調整屬於交易行為，有交易便會有成本，而價格是交易成本的剩餘，在市地重劃開發的過程及後續標得抵費地的土地利用行為中存在交易成本，交易成本的降低可帶來抵費地價格的提升。 重劃完成後各宗抵費地的區位條件與個別條件均不相同，傳統地價影響因素研究多著重於宗地的條件，例如宗地面積、公設可及性、使用分區等，對於土地建築利用與產權整合並無太多著墨，本研究從重劃後建築開發的角度切入，探討以公辦重劃方式所留設的抵費地，其宗地個別條件、土地使用管制、區位、規模及相鄰土地整合難易度，對於交易成本及開發績效有何影響，研究影響抵費地標售價格的因素。 本研究蒐集了林口新市鎮第三期市地重劃區抵費地及相關土地資訊，透過是否建築作為條件篩選，並以189筆合併建築之案例為樣本，運用特徵價格模型，以每平方公尺抵費地標售單價作為應變數，選取13個可能影響抵費地標售價格的應變數，進行實證分析研究。實證結果顯示抵費地的容積率、所面臨最大路寬、面臨主要道路面寬、至學校距離及標售時間等5個因素對於抵費地標售價格呈現正向影響；而抵費地的面積、鄰地權屬複雜度、深度、寬深比等4個因素對於抵費地標售價格呈現負向影響。 公辦市地重劃政府具有土地分配的規劃及主導權，在抵費地面積有限的情況下，倘能了解抵費地的區位、規模與效益之關係，在訂定分配原則及抵費地劃設時考量上述影響抵費地標售價格的因素，以降低交易成本，即可望降低地主共同負擔或創造盈餘以留供重劃區內建設與管理維護，提升開發效益。 / A urban land readjustment (ULR) project follows the trend of urban development and the urban planning, it readjusts all fragmental and irregular land in the area. After building the land of public in the area, the rest of buildable land will be allocated to landlords and reserve the cost equivalent land by distribution principle. The final purpose of ULR is to adjust the property of land. That way of adjustment is a transaction which comes with the cost, and the residual of transaction cost is price. There is transaction cost in the process of ULR project and in the use of cost equivalent land. The reduction of the transaction cost will rise the price of cost equivalent land. The area and individual condition of all the cost equivalent land is different after URL project. The traditional research of land price focus on the condition of land, for example, the area of the land, the accessibility of public facilities and the land use zoning , but less attention in the using of building and the conformity of property. In this research, we discuss the cost equivalent land with the point of building development by ULR, and the effect of the transaction cost to develop performance by analyzing individual condition, land using constrain, area, size, conformity of land. We also study on the factor of the price of cost equivalent land. In this research, we gather the cost equivalent land of Lin-Kou New Town URL and other related information of land. We use the data screened by purpose to build to integrate 189 lands for the base of experimental research, and using the Hedonic Price Method to set the model in which the unit price of cost equivalent land (per square meter ) as the dependent variable and thirteen impact factors as independent variables.。The experiment result shows that five factors including the floor area ratio of cost equivalent land, the maximum width of neighbor road, the width of neighbor, the distance from school and the time for sale have positive impact to the price of the cost equivalent land. On the other hand, the four factors including the area of the cost equivalent land, the complexity of neighbor property, the depth of land and the ratio of width to depth have negative impact to the price of the cost equivalent land. The government has dominant and plan of land distribution by URL-led by the government, for a limited area of cost equivalent land, if we can realize the relationship among region, size and efficiency of cost equivalent land which effect the price of cost equivalent land, it can reduce the transaction cost when we set up the rule of land and cost equivalent land distribution. We can reduce average burden of landlords or make surplus for construction and management for URL, and also promote the efficiency of development.

近年來影城結合商場、娛樂及餐飲，如雨後春筍般出現，建商售屋亦常以影城為吸睛廣告，對消費者而言，影城對周邊住宅價格影響是否存有關聯性係其購屋選擇關心條件之一，然而現今影城的類型大不相同，且觀察到影城大多座落於商圈內或是與百貨商場結合。因此，本研究將影城、商圈與住宅價格的相關程度做交叉分析，以初步了解其關聯程度，並應用集群分析以控制異質樣本，使得形成同質性的房屋屬性來看影城效果，再透過複迴歸模型分析，探討不同類型的影城、商圈與周邊住宅價格之影響關係。 透過複迴歸模型實證結果得知，影城對台北市周邊住宅價格是有影響的，以總價1,000萬元房屋，平均而言，正向的價差介於41萬元至310萬元之間，負向的價差則介於30萬元至271萬元之間，並且正向的價差高於負向的價差。進一步將影城依經營模式及服務方式分類進行實證，採連鎖經營模式的影城周邊房價價差約為40萬元，而提供複合式服務或僅提供單一式服務的影城，對房價的影響差別不大。該結果可提供消費者在評估購買房屋時之參考，亦可作為開發商與銷售業者在預售屋訂價策略及廣告遵循原則。 此外，由於消費型態的改變，消費者習慣將看電影結合其他休閒活動，這些由百貨公司以及電影院異業結盟的商圈更得消費者青睞，影城進駐與商圈發展之關係相輔相成，進而影響新北市周邊住宅價格，以總價1,000萬元房屋，平均來說，影城進駐商圈後能提升房價139萬元。 / These years combination-area of cinemas, malls, recreation and catering are springing up as well as being the eye-catching advertisement of building-contractors. In terms of house-buyers, the correlation between cinemas and the neighborhood housing price is one of the conditions they care. Nevertheless, with the variety of the cineplex-cinema currently and mostly locate in business districts and department stores, our study make a cross analysis between cineplex, cinema, business districts with residential price for their correlation. With cluster analysis to control the heterogeneous samples and to evaluate the price-effect of cineplex-cinemas under the homogeneous housing. Furthermore to multiple regression analyze cinema stationed-in commercial-district and their neighborhood housing price. Our study confirmed the cineplex-cinemas are influential to the Taipei City surrounding-area residential price. For the ten million house, averagely the positive-impact is between four hundred ten thousand to 3.1 million. Negative-impact is between three hundred thousand to 2.71 million. Besides, the housing price different positive effect is higher than the negative one. Further to verify type-mode cinemas: the price difference in neighborhood-area is four hundred thousand with franchise management type-mode. However, there is no price difference with complex and single-service type-mode. Housing-buyer can take the result as consideration during purchasing houses as well as real estate developers and salesman in pricing strategy and advertisement principles of pre-sale houses. Additionally, with the change of consumption patterns, consumers get use to watch movies together with other recreation. Thus commercial-districts combining with department stores and cinemas are more favored. The cinema stationed-in and commercial-district development are complemented each other, therefore to affect the neighborhood housing price in New Taipei City. For the ten million house, averagely, the positive price-different effect is one million and three hundred nighty thousand after the cinema stationed-in.

人類的思考決策與行為結果，大多是依循一個明確的市場制度而為的，故市場經濟制度的重要性，就在於有利個人形成如何有效配置資源的合理預期，同時也富有約束及激勵個人行為的誘因，讓個人能在制度約束的條件下，追求自利動機的極大化。都市空地再利用及市地重劃的法令制度，就具有這樣的經濟效應，所以本研究主要在探討，在空地再利用及市地重劃制度的約束與激勵條件下，都市中的經濟人行為決策與個體互動所形成的空間外部性，其會如何反應在住宅土地價格的變動，以探討空地再利用及空地開發所產生的空間外部性對住宅土地價格之影響。本文主要包括（1）空地再利用對住宅土地價格之影響；（2）空地再利用及鄰里土地開發對新建住宅土地價格之影響；（3）空間外部性、交易成本與市地重劃對新建住宅土地價格之影響等。 首先，有關「空地再利用對住宅土地價格之影響」，本文以Geoghegan（2002）的理論模型為基礎，主要探討空地再利用所產生的開放空間效果，其對於住宅土地價格的影響分析。本文運用地理資訊系統及空間計量分析，並建立民國九十三年及九十四年住宅土地的特徵價格模型，研究對象以台南市為例。實證結果發現，公有空地再利用作為公園及臨時停車場使用，其對於住宅土地價格產生正的空間外溢效果，此表示空地再利用所產生的開放空間的外溢利益，其大於空地管理維護的外溢成本。私有空地再利用作為運動場使用，民國九十三年的外部成本大於外部利益，但於民國九十四年的外部利益大於外部成本，故就私有空地再利用的外溢效果而言，私有空地再利用作為運動場使用的外溢效果可由負向轉為正向。公有空地再利用的邊際價格較私有空地再利用的邊際價格為高，故政府應繼續推動公有空地再利用的政策，以彌補鄰里的開放空間不足。 其次，有關「空地再利用及鄰里土地開發對新建住宅土地價格之影響」，本文主要在探討空地再利用所產生的開放空間效果，及鄰里土地開發所產生的空間外溢效果，兩者對於新建住宅土地價格的影響分析，本研究範圍以台南市民國九十三年的新建住宅為對象。實證結果發現，公有空地再利用對於新建住宅土地價格具有正的空間外溢效果，此表示開放空間的外溢利益大於空地管理維護的外溢成本。鄰里土地開發作為住宅使用時，會對新建住宅土地價格形成正的空間外溢效果，但當鄰里土地開發作為工廠使用時，卻會對新建住宅土地價格形成負的空間外溢效果。最有價值的住宅土地為，開放空間數量最大化及土地開發數量的最小化，故當永久性的開放空間(如鄰里公園)、可開發性的開放空間(如公有空地再利用)數量增加，以及鄰里工廠土地開發數量減少等，其將有助提高新建住宅土地價格。 最後，有關「空間外部性、交易成本與市地重劃對新建住宅土地價格之影響」，本文主要從政府實施市地重劃有助於降低交易成本的觀點，探討政府實施市地重劃制度及建設商的土地開發行為所產生的空間外部性，其反應在新建住宅土地價格的影響。本文運用空間計量分析及地理資訊系統，並結合Box-Cox函數轉換，以建構具有空間外部性的特徵價格模型。本文研究範圍以台南市民國八十年、八十五年及九十年的新建住宅為對象。實證結果得知，空間誤差Box-Cox模型對於空間外部性提供良好的解釋力。政府實施市地重劃之後對住宅土地價格具有正向外部性，且有政府參與的重劃區所反應的住宅土地邊際價格為正向關係。而建設商個體互動關係則具有正向及負向外部性，並在長期下出現由正向轉為負向之情況，此證明空間外部性會呈現相互性的現象，進而反應在新建住宅土地價格上會有增減的作用。 / This study mainly assumes that development of vacant land give rise to spatial externalities from the reuse of vacant land and the institution of land readjustment. Finally, this effect will be reflected in the residential land prices. To test this hypothesis, this study applies spatial econometric and geographic information systems based on hedonic pricing model. This study mainly includes that (1) The Impact of Reusing Vacant Land on Residential Land Prices, ( 2) The Impact of Reusing Vacant Land and Neighboring Land Development on Newly-Built Residential Land Prices, ( 3) The Impact of Spatial Externalities, Transaction Costs and Land Readjustment on Newly-Built Residential Land Prices, and evidence from Tainan City in Taiwan. First of all, about’ The Impact of Reusing Vacant Land on Residential Land Prices’, This study is based on Geoghegan’s theoretical model (2002) and focuses on open space effect of reusing vacant land on residential land prices. The data are selected from residential land prices in 2004, 2005. Empirically, I find that the reuse of publicly-owned vacant land used for green and park has positive spatial spillover effects. This means that spillover benefit of open space is more than spillover cost of vacant land management. The reuse of privately-owned vacant land used for sport has negative spatial spillover effects in 2004. The marginal price of publicly-owned vacant land is over than privately-owned vacant land. It is concluded that, the government should continue promoting the reuse of publicly-owned vacant land and used for open space of the neighborhood. Secondly, about’ The Impact of Reusing Vacant Land and Neighboring Land Development on Newly-Built Residential Land Prices’, This study focuses on open space effect of reusing vacant land and spatial spillover effect of neighboring land development on newly-built residential land prices. The data are selected from newly-built residence in 2004. Empirically, I find that the reuse of publicly-owned vacant land has positive spatial spillover effects on newly-built residential land prices. This means that spillover benefit of open space is more than spillover cost of vacant land management. Land development used for residence has positive spatial spillover effects and used for factory has negative spatial spillover effects in a neighborhood. This means that more the reuse of publicly-owned vacant land will increase newly-built residential land prices, while more land development used for factory will decrease newly-built residential land prices in a neighborhood. Finally, about’ The Impact of Spatial Externalities, Transaction Costs and Land Readjustment on Newly-Built Residential Land Prices’, This Study mainly assumes that land readjustment can reduce transaction costs in terms of the spatial externalities from developers’ behavior and the institution of land readjustment. This effect will be reflected in the newly-built residential land prices. To test this hypothesis, I apply spatial econometric analysis and geographic information systems based on the Box-Cox hedonic pricing model. The data are selected from newly-built residence from between 1991, 1996 and 2001. Empirically, I find that the spatial error Box-Cox model is appropriate for engaging in spatial externalities analysis. This is because the results show that land readjustment gives rise to positive spatial externalities on newly-built residential land prices and the marginal price of residential land in areas where land readjustment takes place is also positive. The effects of developers’ interactions give rise to positive spatial externalities but these change to negative spatial externalities in relation to newly-built residential land prices in the long term. It is concluded that, land readjustment as implemented by the government really does contribute to reducing uncertainty in the land development process, and also reduces transaction costs between construction practitioners and landlords.

估價於房地產市場實扮演著一不可或缺的角色，精確的估價不僅可提供消費者正確極充分的購屋資訊，亦為政府擬定政策方針之基礎。由於台灣房地產市場為一不完全市場，消費者在購屋的同時更常因資訊的不健全而遭受不必要之損失，因此精確及流通之估價資訊實為健全台灣房地產市場之首務。 鑑於過去估價技術仍未成熟，所佔之房價常無法令人信服。本研究欲以類神經網路之功能，將其原理應用於房地產估價上，試圖解決過去估價方法本身之缺失，並作為估價人員輔助之工具。本研究主要以倒傳遞及理解倒傳遞類神經網路與特徵價格法進行公證比較分析，並以特徵價理論為基礎，利用類神經網路得出影率房地產價格更具代表性之因素，以做為未來建立房地產估價輔助系統之參考。 為了解不同的資料型態是否會使類神經網路有不同的學習效果，本研究將資料分為四組實驗設計，分別對不同的資料型態進行測試，研究結果顯示類神經網路對於資料型態較為敏感，其中又以理解倒傳遞類神經網路為最，使得其在預測能力上易受異常點或極端值的影響，而有好壞差異較大的情況。即類神經網路之學習效果端視資料是否具代表性而定。

國內近年來有許多文獻在進行特徵價格模型預測時，避免樣本中存在異常點會造成模型估計值產生偏差，會使用統計軟體進行異常點檢測，但皆是直接將檢測出的異常點刪除，未加以著墨探究這些異常點的特徵結構、成因及特色等。因此，本研究透過統計檢定方法，探討刪除異常點前後整體樣本的特徵結構變化，並以個別估價觀點加以探討住宅交易樣本異常點的成因與特色，藉此歸納出實價登錄資料未揭露的重要特徵，以及迴歸模型搜尋疑似申報不實案件之可行性。 透過敘述統計及樣本結構差異檢定結果發現，異常樣本的離散程度相對原始樣本與正常樣本較大，且經過刪除異常點的正常樣本特徵結構差異程度縮小；異常點的形成可能受到區位變數無法反映實際情況及樣本群聚程度影響，也可能因模型未納入某些重要的特徵變數，而使隱含該變數的樣本被判斷為異常點；異常樣本與正常樣本的成交總價、土地坪數、建物坪數、總樓層、所在樓層及屋齡等變數平均數、變異數及中位數有顯著差異。 藉由個案分析結果歸納，可能因異常個案的住宅屬性存在整幢大樓住商混合使用、特殊鄰居、附屬建物占比過高、高總價豪宅產品、都更效益、增建效益、裝潢效益、約定專用空間效益、樓層高度挑高、獨特視野景觀或特殊區位條件；外部環境存在鄰近嫌惡設施或迎毗設施；交易情況存在買方身分特殊之影響，但受限於實價登錄未要求登載並揭露這些特徵，故模型未考量這些因素對價格的影響，使得模型可能將隱含這些特徵的樣本判斷為異常點，並進而影響模型預測結果。另外也發現，實價登錄資料存在登載錯誤及價格申報不實的情況，且可能被模型判斷為異常點。 / Many literatures use statistics-way to detect outliers in preventing any extreme deviation in hedonic price model prediction. Nevertheless, deleting the outliers instead of investigation into the structures, causes and features. Hence, this thesis studies the feature structures variation of the sample before and after deleting the outliers and with the valuations by appraisers’ perspective to inquire into the factors and features of the outliers in residential transactions. Thereby to summarize the significant features that are not disclosed by real price registration and feasibility in searching the possible false declaration of price by regression. Through descriptive statistics and sample structural difference parametric and nonparametric test shows the discreteness level of singular (outliers only) samples is greater than the primary (outliers including) and normal (outliers deleting) samples and the feature structure variation lessened after deleting the outliers in normal samples. The formation of outliers may be influenced by location variable not able to reflect actual circumstances and level of clustering in samples. Maybe some significant variables are not subsumed into the model, which leads to the judgement of samples with this variable to be outliers. The mean, variance and median in total traded price, land size, building size, total floors, exact floor and house age of singular samples are notably different with normal ones. With the analysis of cases, the possible reasons may be residential and commercial mixed-use in building, peculiar neighbors, high proportion of accessory building, luxury houses, urban renewal benefits, building addition benefits, interior decoration benefits, agreed space benefits, high-ceiling benefits, unique view or location, YIMBY and NIMBY property in environment and special relationship between the buyer and seller. Nevertheless, due to the nondisclosure of these features in real price registration that the model does not take these into consideration. That leads to the judgement of samples with these features as outliers and affects the model prediction. Also the registration error and false declaration in price may also be judged as outliers.