https://github.com/theairbend3r/mice-memory-response
Effect of memory on current response in mice using methods from computational neuroscience and machine learning.
https://github.com/theairbend3r/mice-memory-response
computational-neuroscience data-analysis data-science machine-learning neuroscience python
Last synced: 10 days ago
JSON representation
Effect of memory on current response in mice using methods from computational neuroscience and machine learning.
- Host: GitHub
- URL: https://github.com/theairbend3r/mice-memory-response
- Owner: theairbend3r
- License: gpl-3.0
- Created: 2021-07-14T16:04:04.000Z (almost 5 years ago)
- Default Branch: main
- Last Pushed: 2026-05-10T20:19:40.000Z (about 1 month ago)
- Last Synced: 2026-05-10T22:23:26.576Z (about 1 month ago)
- Topics: computational-neuroscience, data-analysis, data-science, machine-learning, neuroscience, python
- Language: Python
- Homepage:
- Size: 17.7 MB
- Stars: 1
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# Mice Memory Response
[](https://spiking-brains.readthedocs.io/en/latest/?badge=latest)
Study the effect of memory on current response (to behavioral tasks) in mice using methods from computational neuroscience and machine learning.
## Installation
### Clone the repository.
```sh
git clone https://github.com/theairbend3r/spiking-brains.git
```
### Install the packages.
Using Conda.
```sh
conda env create -f spiking-brains.yml
```
Using Pip.
```sh
pip install requirements.txt
```
## Content
The modules reside in the package `./app`.
Following are the notebooks that use function from `./app/` to perform analysis.
1. Exploratory Analysis
2. Behaviour Analysis
3. Neurons Analysis
4. Phenomena Analysis
5. Machine Learning Modelling
## Experiment and Analysis
### Goal
- Study the effect of memory on current response (to behavioral tasks) in mice using machine learning.
### Hypothesis
- Previous responses to visual stimulus, by the mouse, may affect its present response.
### Dataset
- A subset of the Steinmetz dataset (Steinmetz et al, 2019).
- It contains 39 sessions from 10 mice.
- The mice were shown 2 images and had to determine which image had the highest contrast.
### Method
- Train a logistic regression model to predict the mouse's response given the following input variables for `current timestamp - 1`.
- Feedback type
- Feedback time
- Reward time
- Response type
- Contrast left
- Contrast right
- Tune the model and use 8-fold cross validation to gauge accuracy.
- Plot the confusion matrix to compare the actual mouse response vs the model's response.
- Analyse the beta-weights per input variable to see its effect on the response.
### Results
- See `05_modelling.ipynb` notebook.
## Meta
Akshaj Verma – [@theairbend3r](https://twitter.com/theairbend3r).
Distributed under the GNU GPL-V3 license. See `LICENSE` for more information.
[https://github.com/theairbend3r/spiking-brains](https://github.com/theairbend3r/spiking-brains)