A novel neural network technique for the valuation of residential property

Lewis, Owen M.; Ware, Andrew; Jenkins, David H.

doi:10.1007/bf01424227

Cited by 22 publications

(7 citation statements)

References 2 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The SOM network model has been applied in over 5,384 applications in numerous different areas (Oja, Kaski, and Kohonen, 2003). Kaski, Sinkkonen, and Peltonen (2001) and Charalambous, Charitou, and Kaourou (2000) used SOM to predict bankruptcies, and Lewis, Ware, and Jenkins (1997) used it for property valuation. Moreover, Serrano‐Cinca (1996, 1997) used SOM for financial diagnosis and the classification of financial information, respectively.…”

Section: Literature Reviewmentioning

confidence: 99%

Integrating Neural Networks for Risk‐Adjustment Models

Hsu¹,

Lin²,

Yang³

2008

J of Risk & Insurance

View full text Add to dashboard Cite

This article demonstrates the possibility of an alternative approach for risk-adjustment models. In the proposed model the risk characteristics of the beneficiary's health within the same cohort classified by Self-Organizing Map network are highly homogeneous, whereas the numbers of individuals within each cohort remain sufficient to allow further investigation of the causal effect from clustered data. A comparison of different models by the 10-fold cross-validation reveals that the performance improvement in the proposed integration model is both significant and stable across the estimation and validation sampling. Copyright (c) The Journal of Risk and Insurance, 2008.

show abstract

Section: Literature Reviewmentioning

confidence: 99%

Integrating Neural Networks for Risk‐Adjustment Models

Hsu¹,

Lin²,

Yang³

2008

J of Risk & Insurance

View full text Add to dashboard Cite

show abstract

“…First, a Kohonen SOM was used to cluster Census data, and secondly, a genetic algorithm approach was investigated. (Lewis et al, 1997). Census aggregates were passed as input vectors to a Kohonen SOM that clusters data according to their cross-characteristic similarities.…”

Section: Discussionmentioning

confidence: 99%

“…The appeal of this approach is in its ability to estimate the trainability of each cluster using the Gamma test [9] (see Lewis et al [1] for a complete description of the process). This permits only 'useful' clusters to form training sets for MLP networks.…”

Section: Kohonen Self-organising Map Approachmentioning

confidence: 99%

“…Two sub-optimal methods are considered in this paper. The first, which is described more fully elsewhere [1], uses a Kohonen SelfOrganising Map to cluster similar characteristics, followed by an investigation of the usefulness of these clusters as training sets for an MLP (Section 4). The second approach uses Genetic Algorithms to generate random stratas that are refined in successive generations (Section 6).…”

Section: Practical Interpretationmentioning

confidence: 99%

See 1 more Smart Citation

Identification of Residential Property Sub-Markets using Evolutionary and Neural Computing Techniques

Lewis

Ware

Jenkins

2001

Neural Computing & Applications

View full text Add to dashboard Cite

This paper expands on previous work considering methods of stratifying property data in order to enhance its susceptibility to modelling for mortgage value estimation. Previous work [1] considered a clustering approach using a Kohonen Self-Organising Map (SOM) to stratify the training data prior to training a suite of MLPs. Although the results were encouraging, the approach suffers from its estimation of trainability post-clustering. The following method ameliorates the approach by replacing the static clustering step with a dynamic genetic algorithm implementation. The results show a healthy improvement in accuracy over the non-stratified approach, and a more consistent level of accuracy compared with the Kohonen SOM approach. The paper concludes by analysing the underlying content of the derived stratas, thus providing a 'human readable' element to the approach that enhances its potential for acceptance by valuation institutions for as a complementary technique to traditional valuation methods.

show abstract

“…The sales comparison approach is one of most widely used valuation methods, involving hedonic models adjusted by multiple regression analysis (MRA) to deal with differences between the units (Cesare, 1998; Cuthbertson et al , 1992; Iakson, 1998; Mark and Goldberg, 1988; Neter et al , 1990; Ward, 2001). Nowadays, MRA is widely used to develop hedonic mass appraisal models but in recent years several authors have presented neural network applications in property valuation (Cechin et al , 2000; Lewis et al , 1997; McCluskey, 1996; McCluskey and Borst, 1997; McCluskey et al , 2000; McGreal et al , 1998; Nguyen and Cripps, 2001; Rossini, 1997).…”

Section: Related Work – Location In Camamentioning

confidence: 99%

A new approach to spatial analysis in CAMA

González

Soibelman

Formoso

2005

Property Management

View full text Add to dashboard Cite

PurposeAvailable literature claims that location is a key attribute in the housing market. However, the impact of this attribute is difficult to measure and the traditional hedonic approach using subjective assessments is problematic. This paper seeks to explore trend surface analysis technique, attempting to provide an alternative way to measure location values.Design/methodology/approachTSA works in a similar way to other response surface methods but it is implemented directly in regression models, using a set of combinations of the co‐ordinates of properties in several power degrees. It can also be implemented in artificial neural networks, taking advantage of the neural ability in non‐linear domains. This work presents a comparison between traditional regression approach, error modelling, response surfaces, and TSA. ANN is also used to estimate some models, comparing their results. The objective is to verify the behaviour of TSA in hedonic models. A case study was carried using data of over 30,000 sales tax data of apartments sold in Porto Alegre, a southern Brazilian town.FindingsThe results indicates that TSA is an effective tool for the spatial analysis of real estate, because TSA models are similar to other approaches, but are developed with less expert work.Originality/valueThis paper presents an application of TSA in real estate market, which is an interesting alternative to traditional measures of location attributes.

show abstract

A novel neural network technique for the valuation of residential property

Cited by 22 publications

References 2 publications

Integrating Neural Networks for Risk‐Adjustment Models

Integrating Neural Networks for Risk‐Adjustment Models

Identification of Residential Property Sub-Markets using Evolutionary and Neural Computing Techniques

A new approach to spatial analysis in CAMA

Contact Info

Product

Resources

About