My current research incorporates temporal convolutional networks, evolutionary strategies, natural language processing into an architecture that takes into the multi-dimensionality of financial news (company relationship and emotional analysis) and financial markets (macro and sector economics) to better predict market fluctuations.
Evolution Strategies (ES) are a sub-class of nature-inspired direct search (and optimization) methods which belong to the evolutionary algorithms (EAs). These use mutation, recombination, and selection applied to a population of individuals containing candidate solution to evolve iteratively better solutions. By optimizing both the network structures and hyper-parameters I look to optimize for both performance and accuracy where the industry heads to larger and more complex neural networks (e.g. large language models). In the next section I will cover some relevant background and a recent article of interest related to the motivation of this ongoing research.
Biography
Mr. McCarthy is an experienced Senior-level IT executive with over 18 years experience whose passion is to inspire growth. His background includes leadership, organizational evolution, M&A, agile and devops change agent, recruiting interns and professionals, developing/coaching teams and individuals, system architecture, system development and design, application programming, product support, business analysis, and professional training.
- Architecture, Risk and Big Data
- PhD Candidate Machine Learning
- Certified Deep Learning (Anomaly Detection, NLP, Recommender systems)
- AI Committee Member @ AnitaB.org
- Graduate Instructor
Anomaly Detection
Natural language processing
Recommendation systems