Resources
Curated datasets, tutorials, tools, and educational materials for machine learning in finance research.
Wiki
Key concepts and methodologies used in our research
Systematic Literature Reviews Explained
A systematic literature review (SLR) is a form of secondary research that uses a clearly defined, replicable methodology to identify, select, critically appraise, and synthesize all relevant primary studies addressing a specific research question. Unlike narrative reviews, which are valuable for expert commentary but vulnerable to selection bias and lack reproducibility, SLRs follow a predefined protocol that makes every methodological decision explicit and auditable.
Five core principles
| Principle | Description |
|---|---|
| Transparency | every step is documented for scrutiny |
| Reproducibility | another researcher can replicate the process |
| Comprehensiveness | aims to identify all relevant studies |
| Explicit methodology | criteria and strategies specified a priori |
| Bias minimization | systematic procedures reduce subjective selection |
Systematic vs. narrative reviews
| Dimension | Narrative | Systematic |
|---|---|---|
| Protocol | typically absent | defined a priori |
| Search | selective, implicit | comprehensive, explicit |
| Study selection | reviewer discretion | predefined criteria |
| Quality appraisal | rarely conducted | standardized |
| Reproducibility | low | high |
Finance-specific challenges
| Challenge | Description |
|---|---|
| Working papers | SSRN/NBER serve as primary channels; versioning complicates deduplication |
| Conference-dominant fields | ML disseminates via NeurIPS, ICML, ArXiv rather than journals |
| Proprietary data | reliance on WRDS, Bloomberg, Refinitiv creates reproducibility barriers |
| Rapid evolution | reviews risk obsolescence; computational pipelines enable living reviews |
Datasets
Open datasets for financial ML research ·
Loading datasets...
Tools & Libraries
Software tools for ML in finance research and development
PyTorch / TensorFlow
Deep learning frameworks for building and training neural networks.
XGBoost / LightGBM
Gradient boosting libraries for high-performance ensemble learning.
Zipline / Backtrader
Backtesting frameworks for testing trading strategies on historical data.
cvxpy
Convex optimization library for portfolio optimization and risk management.
PyPortfolioOpt
Portfolio optimization library with mean-variance, risk parity, and more.
Stable-Baselines3
Reliable implementations of reinforcement learning algorithms in PyTorch.
OpenAlex API
Open catalog of scholarly works for literature review and research discovery.
Academic Resources
Journals, conferences, and academic outlets