Yiannis Kanellopoulos

The ability to explain in understandable terms, why a machine learning model makes a certain prediction is becoming immensely important, as it ensures trust and transparency in the decision process of the model. Complex models, such as ensemble or deep learning models, are hard to interpret. Various methods have been proposed that deal with this matter. Shapley values provide accurate explanations, as they assign each feature an importance value for a particular prediction. However, the… Εμφάνιση περισσότερων

The ability to explain in understandable terms, why a machine learning model makes a certain prediction is becoming immensely important, as it ensures trust and transparency in the decision process of the model. Complex models, such as ensemble or deep learning models, are hard to interpret. Various methods have been proposed that deal with this matter. Shapley values provide accurate explanations, as they assign each feature an importance value for a particular prediction. However, the exponential complexity of their calculation is dealt efficiently only in decision tree-based models. Another method is surrogate models, which emulate a black-box model’s behavior and provide explanations effortlessly, since they are constructed to be inter- pretable. Surrogate models are model-agnostic, but they produce only approximate explanations, which cannot always be trusted. We propose a method that combines these two approaches, so that we can take advantage of the model-agnostic part of the surrogate models, as well as the explanatory power of the Shapley values. We introduce a new metric, Topj Similarity, that measures the similitude of two given explanations, produced by Shapley values, in order to evaluate our work. Finally, we recommend ways on how this method could be improved further. Λιγότερα

Algorithmic systems make decisions that have a great impact in our lives. As our dependency on them is growing so does the need for transparency and holding them accountable. This paper presents a model for evaluating how transparent these systems are by focusing on their algorithmic part as well as the maturity of the organizations that utilize them. We applied this model on a classification algorithm created and utilised by a large financial institution. The results of our analysis indicated… Εμφάνιση περισσότερων

Algorithmic systems make decisions that have a great impact in our lives. As our dependency on them is growing so does the need for transparency and holding them accountable. This paper presents a model for evaluating how transparent these systems are by focusing on their algorithmic part as well as the maturity of the organizations that utilize them. We applied this model on a classification algorithm created and utilised by a large financial institution. The results of our analysis indicated that the organization was only partially in control of their algorithm and they lacked the necessary benchmark to interpret the deducted results and assess the validity of its inferencing. Λιγότερα

In 2016, the Software Improvement Group (SIG) collaborated with publisher O’Reilly Media to survey 1,400 software developers on topics related to code quality. The intent of the survey was to uncover trends in overall attitudes, working assumptions, resource distribution, and individual and team behaviors around code qual‐ ity. In general, the results of the survey reinforced SIG’s findings from prior surveys and years of field work with software teams: code quality is valued in principle yet… Εμφάνιση περισσότερων

In 2016, the Software Improvement Group (SIG) collaborated with publisher O’Reilly Media to survey 1,400 software developers on topics related to code quality. The intent of the survey was to uncover trends in overall attitudes, working assumptions, resource distribution, and individual and team behaviors around code qual‐ ity. In general, the results of the survey reinforced SIG’s findings from prior surveys and years of field work with software teams: code quality is valued in principle yet often measured and managed unevenly—or not at all—in the day-to-day practices of software development organizations. Λιγότερα

Η ελαχιστοποίηση του ρίσκου, η μεγιστοποίηση της σταθερότητας και η υποστήριξη της ανάπτυξης του νέου οργανισμού είναι το τρίπτυχο των προκλήσεων που αντιμετωπίζει η Διεύθυνση Πληροφορικής (ΙΤ) στο πλαίσιο μίας διαδικασίας συγχώνευσης και εξαγοράς (Σ&Ε).

IT management consulting is usually perceived as a technical endeavour in which consultants focus on bits and pieces. However, it primarily has to do with communicating and mobilising people within an organisation to change; the technical things will follow. The guiding principles to achieve this are: thinking about others and not yourself, being factual, communicating on multiple levels, showing empathy, and thinking out of the box. This article presents how these five principles were applied… Εμφάνιση περισσότερων

IT management consulting is usually perceived as a technical endeavour in which consultants focus on bits and pieces. However, it primarily has to do with communicating and mobilising people within an organisation to change; the technical things will follow. The guiding principles to achieve this are: thinking about others and not yourself, being factual, communicating on multiple levels, showing empathy, and thinking out of the box. This article presents how these five principles were applied in different contexts to our software improvement consulting practice, but had the same effect; to inspire people to change. Λιγότερα