28 Aralık 2022 tarihinde Anadolu Üniversitesi Eğitim Fakültesi seminerleri kapsamında gerçekleştirilen "Açıklanabilir Yapay Zeka" sunum materyallerini içerir.

Official implementation of "insertion/deletion metrics-aware explanation-based optimization"

https://github.com/jphall663 -- https://github.com/jphall663/awesome-machine-learning-interpretability -- https://www.youtube.com/watch?v=d6j6bofhj2M&list=PLV8yxwGOxvvovp-j6ztxhF3QcKXT6vORU&index=3

Dataiku XAI showcase project using XAI features like PDP and SHAP

Official code for MAFIA (MAsk-based Feature Interaction Attribution): Explain BERT model's decision in NLI task.

來自《Econometric Analysis》書本的信用卡核發資料，結合XAI 套件SHAP。

code for paper in NLDB'23

Fundamentals of Interpretable Data Science

Code repository for WYM's Demo submitted at VLDB2023 and CIKM2023.

Cardiovasular Disease Detection using Naive Bayes, Logistic Regression, Random Forest & Support Vector Machine, while comparing the Naive Bayes models with the rest. LIME was also used to explain the predictions of the model.

Collection of papers used for a scoping review of explanation types and need indicators in human–agent interact / robotics / human–agent collaborations

Playing by AI's rules - Matching Pairs / Memory with CLIP Cosine Similarity!

Jekyll repository for Toybox webpage.

Visualize, analyze, and explore NLP language models. Ecco creates interactive visualizations directly in Jupyter notebooks explaining the behavior of Transformer-based language models (like GPT2, BERT, RoBERTA, T5, and T0).

설명 가능한 인공지능 끝내기

Descriptive analysis and QSAR modelling for tox_21 datasets

Targeted semantic multimodal input ablation. Official implementation of the ablation method introduced in the paper: "What Vision-Language Models 'See' when they See Scenes"

Bachelor thesis work: Debugging of a counterfactual-explanation library (CFNOW) and its usage in the context of eXplainable Affective Computing. Final evaluation: 30/30.

Code and data for the paper "When Do People Want an Explanation?", presented at HRI'24