https://github.com/mshuqair/rl-based-adaptive-classifier

Deep Reinforcement Learning based adaptive time-series binary classifier.
https://github.com/mshuqair/rl-based-adaptive-classifier

classification deep-reinforcement-learning keras parkinsons-disease python reinforcement-learning

Last synced: about 1 month ago
JSON representation

Deep Reinforcement Learning based adaptive time-series binary classifier.

• Host: GitHub
• URL: https://github.com/mshuqair/rl-based-adaptive-classifier
• Owner: mshuqair
• Created: 2023-01-18T16:12:27.000Z (almost 2 years ago)
• Default Branch: main
• Last Pushed: 2024-08-12T14:30:24.000Z (5 months ago)
• Last Synced: 2024-08-12T16:30:07.783Z (5 months ago)
• Topics: classification, deep-reinforcement-learning, keras, parkinsons-disease, python, reinforcement-learning
• Language: Python
• Homepage:
• Size: 1.11 MB
• Stars: 2
• Watchers: 2
• Forks: 1
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • Citation: CITATION.cff

Awesome Lists containing this project

README

        
# Deep Reinforcement Learning Based Adaptive Classifier

This unique approach uses Reinforcement Learning (RL) to discern shifts in data stream distributions during state transitions.

Training an RL agent to recognize these transitions makes it adept at identifying transitions in new data.

Instead of static models, our agent interacts with the data's dynamics and makes optimal classification decisions.

This RL-driven framework prioritizes understanding changes in data distribution, making it robust against inter and intra-data variations.

![](figures/figure_rl_structure.png)

**Figure 1.** The proposed reinforcement learning-based adaptive classification framework.

![](output/model_output.png)

**Figure 2.** The model prediction for Walking vs. Non-walking (1 vs. 2) in one participant of the MHEALTH dataset.

## Updates

More updates regarding the description are coming soon. I will also update the code to make it more generalizable.

## Deep Reinforcement Learning Adaptive Classification of PD Medication State

- The preliminary results of this project were published at the IEEE ICDM 2022 Conference.

Incremental Learning in Time-series Data using Reinforcement Learning: https://doi.org/10.1109/ICDMW58026.2022.00115

- The IEEE Journal of Biomedical and Health Informatics has recently published an extensive extension of this work.

Reinforcement Learning-Based Adaptive Classification for Medication State Monitoring in Parkinson's Disease: https://doi.org/10.1109/JBHI.2024.3423708

## Code Requirements and Compatability

The code was run and tested using the following:

- Python			3.8.10

- tensorflow		2.8.4

- keras				2.8.0

- h5py				3.10.0

- matplotlib		3.6.3

- numpy				1.20.3

- pandas			1.3.4

- scikit-learn		1.3.2

- scikit-multiflow	0.5.3

## Conclusions