Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Tackling the Accuracy-Interpretability Trade-off: Interpretable Deep Learning Models for Satellite Image-based Real Estate Appraisal

Tackling the Accuracy-Interpretability Trade-off: Interpretable Deep Learning Models for... Deep learning models fuel many modern decision support systems, because they typically provide high predictive performance. Among other domains, deep learning is used in real-estate appraisal, where it allows extending the analysis from hard facts only (e.g., size, age) to also consider more implicit information about the location or appearance of houses in the form of image data. However, one downside of deep learning models is their intransparent mechanic of decision making, which leads to a trade-off between accuracy and interpretability. This limits their applicability for tasks where a justification of the decision is necessary. Therefore, in this article, we first combine different perspectives on interpretability into a multi-dimensional framework for a socio-technical perspective on explainable artificial intelligence. Second, we measure the performance gains of using multi-view deep learning, which leverages additional image data (satellite images) for real estate appraisal. Third, we propose and test a novel post hoc explainability method called Grad-Ram. This modified version of Grad-Cam mitigates the intransparency of convolutional neural networks for predicting continuous outcome variables. With this, we try to reduce the accuracy-interpretability trade-off of multi-view deep learning models. Our proposed network architecture outperforms traditional hedonic regression models by 34% in terms of MAE. Furthermore, we find that the used satellite images are the second most important predictor after square feet in our model and that the network learns interpretable patterns about the neighborhood structure and density. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png ACM Transactions on Management Information Systems (TMIS) Association for Computing Machinery

Tackling the Accuracy-Interpretability Trade-off: Interpretable Deep Learning Models for Satellite Image-based Real Estate Appraisal

Download PDF
24 pages

Loading next page...
 
/lp/association-for-computing-machinery/tackling-the-accuracy-interpretability-trade-off-interpretable-deep-ovuSoPemZG
Publisher
Association for Computing Machinery
Copyright
Copyright © 2023 Association for Computing Machinery.
ISSN
2158-656X
eISSN
2158-6578
DOI
10.1145/3567430
Publisher site
See Article on Publisher Site

Abstract

Deep learning models fuel many modern decision support systems, because they typically provide high predictive performance. Among other domains, deep learning is used in real-estate appraisal, where it allows extending the analysis from hard facts only (e.g., size, age) to also consider more implicit information about the location or appearance of houses in the form of image data. However, one downside of deep learning models is their intransparent mechanic of decision making, which leads to a trade-off between accuracy and interpretability. This limits their applicability for tasks where a justification of the decision is necessary. Therefore, in this article, we first combine different perspectives on interpretability into a multi-dimensional framework for a socio-technical perspective on explainable artificial intelligence. Second, we measure the performance gains of using multi-view deep learning, which leverages additional image data (satellite images) for real estate appraisal. Third, we propose and test a novel post hoc explainability method called Grad-Ram. This modified version of Grad-Cam mitigates the intransparency of convolutional neural networks for predicting continuous outcome variables. With this, we try to reduce the accuracy-interpretability trade-off of multi-view deep learning models. Our proposed network architecture outperforms traditional hedonic regression models by 34% in terms of MAE. Furthermore, we find that the used satellite images are the second most important predictor after square feet in our model and that the network learns interpretable patterns about the neighborhood structure and density.

Journal

ACM Transactions on Management Information Systems (TMIS) – Association for Computing Machinery

Published: Jan 16, 2023

Keywords: Interpretability

References

  • Peeking inside the black-box: A survey on explainable artificial intelligence (XAI)
    [,
  • Sanity checks for saliency maps
    [,
  • Towards robust interpretability with self-explaining neural networks
    [,
  • Sam: The sensitivity of attribution methods to hyperparameters
    [,
  • Explainable artificial intelligence (XAI): Concepts, taxonomies, opportunities and challenges toward responsible AI
    [,
  • Beyond spatial auto-regressive models: Predicting housing prices with satellite imagery
    [,
  • Quantifying curb appeal
    [,
  • Random forests
    [,
  • Real estate appraisal residential building 2018
    [,
  • Techniques for interpretable machine learning
    [,
  • Real estate image-based appraisal using mask region based convolutional networks
    [,
  • Greedy function approximation: A gradient boosting machine
    [,
  • Peeking inside the black box: Visualizing statistical learning with plots of individual conditional expectation
    [,
  • Get a static map
    [,
  • Explanations from intelligent systems: Theoretical foundations and implications for practice
    [,
  • Positioning and presenting design science research for maximum impact
    [,
  • A survey of methods for explaining black box models
    [,
  • XAI—Explainable artificial intelligence
    [,
  • Delving deep into rectifiers: Surpassing human-level performance on imagenet classification
    [,
  • Boosting the predictive accuracy of urban hedonic house price models through airborne laser scanning
    [,
  • Combining satellite imagery and machine learning to predict poverty
    [,
  • The right to explanation, explained
    [,
  • Big data in real estate? From manual appraisal to automated valuation
    [,
  • What image features boost housing market predictions? IEEE Trans
    [,
  • Quantifying the impact of location data in real estate appraisal—A gis-based deep learning approach
    [,
  • A comparison of multi-view learning strategies for satellite image-based real estate appraisal
    [,
  • A new approach to consumer theory
    [,
  • An unsupervised approach to geographical knowledge discovery using street level and street network images
    [,
  • Take a look around: Using street view and satellite images to estimate house prices
    [,
  • A survey of multi-view representation learning
    [,
  • Measuring structural, location and environmental effects: A hedonic analysis of housing market in Wroclaw, Poland
    [,
  • House price prediction: Hedonic price model vs
    [,
  • The mythos of model interpretability
    [,
  • A unified approach to interpreting model predictions
    [,
  • Explainable machine-learning predictions for the prevention of hypoxaemia during surgery
    [,
  • Explaining data-driven document classifications
    [,
  • Interactive application of computer assisted mass appraisal and geographic information systems
    [,
  • Get a static map
    [,
  • Interpretable Machine Learning
    [,
  • One picture is worth a thousand words? The pricing power of images in e-Commerce
    [,
  • Sustainable urban regeneration: GIS and hedonic pricing method in determining the value of green space in housing area
    [,
  • Vision-based real estate price estimation
    [,
  • Ensemble manifold regularized multi-modal graph convolutional network for cognitive ability prediction
    [,
  • “Why should I trust you?”: Explaining the predictions of any classifier
    [,
  • Hedonic prices and implicit markets: Product differentiation in pure competition
    [,
  • Stop explaining black box machine learning models for high stakes decisions and use interpretable models instead
    [,
  • Evaluating the visualization of what a deep neural network has learned
    [,
  • Grad-CAM: Visual explanations from deep networks via gradient-based localization
    [,
  • Predicting economic development using geolocated Wikipedia articles
    [,
  • Crop yields prediction with CNN and visual explanation with Grad-RAM
    [,
  • Integrating floor plans into hedonic models for rent price appraisal
    [,
  • Rethinking the inception architecture for computer vision
    [,
  • Sanity checks for saliency metrics
    [,
  • Diabetic retinopathy detection via deep convolutional networks for discriminative localization and visual explanation
    [,
  • CBAM: Convolutional block attention module
    [,
  • A survey on multi-view learning
    [,
  • AI blue book: Vehicle price prediction using visual features
    [,
  • Learning deep features for discriminative localization
    [,