dv-rastogi / ml-pg-project

Stock market prediction using Neural ODEs and GANs

Graduate Machine Learning Project

Modeling Stock Market Returns using Deep Learning

Dataset

CNNPred Dataset

Baselines

1. Arima
2. PCA + DNN (https://jfin-swufe.springeropen.com/articles/10.1186/s40854-019-0138-0)
3. CNNPred (https://www.sciencedirect.com/science/article/abs/pii/S0957417419301915)
4. CNN-LSTM (https://www.hindawi.com/journals/complexity/2020/6622927/)

Final Models

Our primary model is a NeuralODE which outperforms the SOTA CNNPred Baseline.

Our experiments also include generative modeling with NeuralODEs and GANs.

Contributors

1. Devansh Gupta
2. Divyansh Rastogi
3. Rupanshu Yadav