June 2021 - Present
COMPUTATIONAL NEUROSCIENTIST | ML / AI / QUANT
Non-Compete period up to the end of 2022, Florida, USA - As a Data Researcher, I mimic Computational Neuroscience hypotheses and apply scientific / engineering methods to present new solutions for cracking challenging AI/ML/Quant tasks.
- Aiming to solve diverse Quantitative tasks focusing on delivering advanced algorithmic trading strategy for financial markets. Including data mining, asset allocation, risk management, robustness validation, etc.
- Presented a prototype of Brain-Like AI with the key features of fast learning, resistance to overoptimization, computational efficiency, supporting huge models, continuous learning, reinforcement learning, and others.
- Powered by a novel developed bio-neuron and a model of cortical microcolumn as a basic computation unit allowed solving some of the complicated problems of ML/AI easily. The flexible and changing topology of bio-neural networks, sparse connectivity, and vastly parallel computation are a few notable advancements.
- Explored a simple and universal learning rule for bio-neuron, updating weights and mechanism of growing or running dendrites. Moreover, the Brain-Like learning rule for network parts was discovered using L1 of microcolumn and apical dendrites.
- Model of artificial Basal Ganglia (for decision making), Hippocampus (data coding and storing), and Thalamus (information flow management), as a core part of novel AI system, which makes it possible to solve very challenging tasks in a completely new way.
- Revealed technological insights highlighted a way for AGI's practical implementation in a very appropriate environment of financial markets.
(
Read More )