Senior data scientist who teaches machines for a living. I am passionate about implementing creative solutions for never-before-solved/seen problems. I love to automate & optimise processes and work with non-traditional ideas.
Current Interests:
- Artificial Intelligence
- Data Science
- Deep Learning
- Reinforcement Learning
- Optimisation
- Automated Algorithm Selection
- Automated Algorithm Generation
- Fake News Combat
Publications:
- Automated Algorithm Selection: from Feature-Based to Feature-Free Approaches.
- Parkinson’s Disease Diagnosis using Convolutional Neural Networks and Figure-copying Tasks.
- A Neural Approach to Generation of Constructive Heuristics.
- Algorithm Selection Using Deep Learning Without Feature Extraction
- A Deep Learning Approach to Predicting Solutions in Streaming Optimisation Domains
- TRUSTD: Combat Fake Content using Blockchain and Collective Signature Technologies
- Trust-based Ecosystem to Combat Fake News
Projects:
- Customer segmentation using Spark and Pandas (dataset: +500K rows of transactions).
- Customer churn prediction using machine learning with short and long processing pipelines.
- Predicting developer salary using Spark.
- A Tableau dashboard of StackOverflow 2020 Survey.
- Developing a recommendation system using custom Restricted Boltzmann Machine as a subclass-Tensorflow 2.
- Cars CO2 emission prediction.
- Predicting the outcome of the cardiotocogram exam for fetuses.
- Parkinson’s disease diagnosis using Deep Learning (CNN and RNN-LSTM).
- Sentiment analysis using Machine Learning for open domain conversational agent.
- Interactive dashboard for a quality assessment of UK universities’ research.
- Developing an automated extract-transform-load data pipeline to transform a relational database (MySQL) into NoSQL database using Open Flights Dataset (+67,000 records).
- Emotion recognition using Machine Learning.