The aim of this project is to compare Neural Ordinary Differential Equations (NODE) to other neural networks and machine learning algorithms. This was done as part of the summer internship for PES Innovation Lab (May 2020 - August 2020).
This was done in order to compare the performance of NODE when predicting COVID-19 cases and deaths.
The Neural Networks included:
- LSTMs
- GRUs
- RNNs
- Simple neural networks
The Machine Learning algorithms included:
- Gradient Boosting
- Linear Regression
- Polynomial Regression
This was done in order to compare the performance of NODE with respect to the MNIST and CIFAR datasets.
NODE was compared to the following:
- CNN (with and without transfer learning)
- RESNET
Additionally, the time-accuracy tradeoff for NODE was also tested.
INTERNS:
- Aanchal Narendran
- Kartika Nair
- Sujith K
MENTOR:
- Saahil Jain