Homayun Afrabandpey

Homayun Afrabandpey, Michael Spranger

We present a human-in-the-loop approach to generate counterfactual (CF) explanations that preserve global and local feasibility constraints. Global feasibility constraints refer to the causal constraints that are necessary for generating actionable CF explanation. Assuming a domain expert with knowledge on unary and binary causal constraints, our approach efficiently employs this knowledge to generate CF explanation by rejecting gradient steps that violate these constraints. Local feasibility constraints encode end-user's constraints for generating desirable CF explanation. We extract these constraints from the end-user of the model and exploit them during CF generation via user-defined distance metric. Through user studies, we demonstrate that incorporating causal constraints during CF generation results in significantly better explanations in terms of feasibility and desirability for participants. Adopting local and global feasibility constraints simultaneously, although improves user satisfaction, does not significantly improve desirability of the participants compared to only incorporating global constraints.

Homayun Afrabandpey, Hamed Rezazadegan Tavakoli

Efficient model distribution is becoming increasingly critical in bandwidth-constrained environments. In this paper, we propose a simple yet effective approach called Progressive Precision Update (P$^2$U) to address this problem. Instead of transmitting the original high-precision model, P$^2$U transmits a lower-bit precision model, coupled with a model update representing the difference between the original high-precision model and the transmitted low precision version. With extensive experiments on various model architectures, ranging from small models ($1 - 6$ million parameters) to a large model (more than $100$ million parameters) and using three different data sets, e.g., chest X-Ray, PASCAL-VOC, and CIFAR-100, we demonstrate that P$^2$U consistently achieves better tradeoff between accuracy, bandwidth usage and latency. Moreover, we show that when bandwidth or startup time is the priority, aggressive quantization (e.g., 4-bit) can be used without severely compromising performance. These results establish P$^2$U as an effective and practical solution for scalable and efficient model distribution in low-resource settings, including federated learning, edge computing, and IoT deployments. Given that P$^2$U complements existing compression techniques and can be implemented alongside any compression method, e.g., sparsification, quantization, pruning, etc., the potential for improvement is even greater.

Homayun Afrabandpey, Tomi Peltola, Samuel Kaski

Learning predictive models from small high-dimensional data sets is a key problem in high-dimensional statistics. Expert knowledge elicitation can help, and a strong line of work focuses on directly eliciting informative prior distributions for parameters. This either requires considerable statistical expertise or is laborious, as the emphasis has been on accuracy and not on efficiency of the process. Another line of work queries about importance of features one at a time, assuming them to be independent and hence missing covariance information. In contrast, we propose eliciting expert knowledge about pairwise feature similarities, to borrow statistical strength in the predictions, and using sequential decision making techniques to minimize the effort of the expert. Empirical results demonstrate improvement in predictive performance on both simulated and real data, in high-dimensional linear regression tasks, where we learn the covariance structure with a Gaussian process, based on sequential elicitation.

Iiris Sundin, Tomi Peltola, Muntasir Mamun Majumder et al.

Predicting the efficacy of a drug for a given individual, using high-dimensional genomic measurements, is at the core of precision medicine. However, identifying features on which to base the predictions remains a challenge, especially when the sample size is small. Incorporating expert knowledge offers a promising alternative to improve a prediction model, but collecting such knowledge is laborious to the expert if the number of candidate features is very large. We introduce a probabilistic model that can incorporate expert feedback about the impact of genomic measurements on the sensitivity of a cancer cell for a given drug. We also present two methods to intelligently collect this feedback from the expert, using experimental design and multi-armed bandit models. In a multiple myeloma blood cancer data set (n=51), expert knowledge decreased the prediction error by 8%. Furthermore, the intelligent approaches can be used to reduce the workload of feedback collection to less than 30% on average compared to a naive approach.

Seppo Virtanen, Homayun Afrabandpey, Samuel Kaski

A main goal of data visualization is to find, from among all the available alternatives, mappings to the 2D/3D display which are relevant to the user. Assuming user interaction data, or other auxiliary data about the items or their relationships, the goal is to identify which aspects in the primary data support the userś input and, equally importantly, which aspects of the userś potentially noisy input have support in the primary data. For solving the problem, we introduce a multi-view embedding in which a latent factorization identifies which aspects in the two data views (primary data and user data) are related and which are specific to only one of them. The factorization is a generative model in which the display is parameterized as a part of the factorization and the other factors explain away the aspects not expressible in a two-dimensional display. Functioning of the model is demonstrated on several data sets.

Javad Salimi Sartakhti, Homayun Afrabandpey, Nasser Ghadiri

Least Squares Twin Support Vector Machine (LST-SVM) has been shown to be an efficient and fast algorithm for binary classification. It combines the operating principles of Least Squares SVM (LS-SVM) and Twin SVM (T-SVM); it constructs two non-parallel hyperplanes (as in T-SVM) by solving two systems of linear equations (as in LS-SVM). Despite its efficiency, LST-SVM is still unable to cope with two features of real-world problems. First, in many real-world applications, labels of samples are not deterministic; they come naturally with their associated membership degrees. Second, samples in real-world applications may not be equally important and their importance degrees affect the classification. In this paper, we propose Fuzzy LST-SVM (FLST-SVM) to deal with these two characteristics of real-world data. Two models are introduced for FLST-SVM: the first model builds up crisp hyperplanes using training samples and their corresponding membership degrees. The second model, on the other hand, constructs fuzzy hyperplanes using training samples and their membership degrees. Numerical evaluation of the proposed method with synthetic and real datasets demonstrate significant improvement in the classification accuracy of FLST-SVM when compared to well-known existing versions of SVM.