v.vichi3/Neural_Network_for_MOID_Prediction
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions
This repository contains the codes used to explore the data generated by NEOPOP, build the neural network, tune the hyperparameters and analyse the results.
15 commits 1 branch 0 tags 256 MiB
  • Jupyter Notebook 99.9%
Find a file
Exact
Vanessa Vichi 4c2bfb1af9 Jupyter notebook used for 1) training the NN, 2) evaluating its performance on the NEOPOP test set and on the real NEOs dataset
2024-05-12 10:24:46 +02:00
NEOPOP_output_files Output files of the NEOPOP run 2024-05-09 09:44:30 +02:00
split_dataset NEOPOP DataFrame split into training, validation, test 2024-05-09 09:47:58 +02:00
trained_models Saved weights of Models 1,2 after a 500-epoch training (for both models) and a 1000-epoch training for Model 1 2024-05-12 10:19:36 +02:00
.gitignore Initial commit 2024-04-27 13:20:42 +00:00
baseline_performance.ipynb Jupyter notebook for evaluating the baseline performance: comparison of various metrics for the baseline model, the linear regression model and the polynomial regression model of degrees 2 and 3 2024-05-09 09:57:48 +02:00
dataframe.csv NEOPOP DataFrame 2024-05-08 18:38:45 +02:00
dataset_splitting.ipynb Jupyter notebook for splitting the NEOPOP dataset into training, validation and test (with checks over the distribution of the various parts) 2024-05-09 09:54:30 +02:00
expl_data_analysis.ipynb Jupyter notebook for preliminary data exploration 2024-05-09 09:50:42 +02:00
kep_to_att.ipynb Conversion from Keplerian elements to attributable elements 2024-05-11 10:58:40 +02:00
LICENSE Initial commit 2024-04-27 13:20:42 +00:00
neopop_attr.csv DataFrames with the attributable elements 2024-05-12 10:22:38 +02:00
neos_attr.csv DataFrames with the attributable elements 2024-05-12 10:22:38 +02:00
neos_dataframe.csv Pre-processed NEOs DataFrame 2024-05-12 10:20:59 +02:00
neos_dataset_preprocessing.ipynb Jupyter Notebook for pre-processing of the NEOs DataFrame 2024-05-11 10:49:36 +02:00
NeuralNetwork.ipynb Jupyter notebook used for 1) training the NN, 2) evaluating its performance on the NEOPOP test set and on the real NEOs dataset 2024-05-12 10:24:46 +02:00
README.md Initial commit 2024-04-27 13:20:42 +00:00
sbdb_query_results.csv NEOs DataFrame 2024-05-08 18:39:55 +02:00
weight_initialization.ipynb Jupyter notebook for choosing the best initialization technique 2024-05-09 10:04:46 +02:00

README.md

Neural_Network_for_MOID_Prediction

This repository contains the codes used to explore the data generated by NEOPOP, build the neural network, tune the hyperparameters and analyse the results.