ML-notebooks

Agriculture products such as Wheat, rice paddy Minimum support price per Kg. (MSP) forecast.
Artifial neural network
Time series
Frauds detection There is not much data preprocessing involved beacuse data is cleaned and thers is no missing value or outliers. Algorithms are SVM (it is very slow), Decision Tree and Light GBM. Highest accuracy on test data is above 99 %, metrics are roc curve, rmse score and auc.
Pandas and Numpy Some lessons from book pandas data wrangling by wes mckinney
Quora duplicate question It is binary calssfication problem with accuracy above 80%
Common literature target It is regression problem and highest rmse score of 0.749. It is about predicting target score ( -1<=score<=1 ) of small paragraph (which is a small story or meaningful content)

Name		Name	Last commit message	Last commit date
Latest commit History 12 Commits
.gitignore		.gitignore
Artificial_Neuron_Network.ipynb		Artificial_Neuron_Network.ipynb
Bank_Transaction_Frauds.ipynb		Bank_Transaction_Frauds.ipynb
Ch-10 Data Aggregation and Group.ipynb		Ch-10 Data Aggregation and Group.ipynb
Ch-11 Time-Series-- important chapter .ipynb		Ch-11 Time-Series-- important chapter .ipynb
MSP_price_pred.ipynb		MSP_price_pred.ipynb
Numpy_and_copy.ipynb		Numpy_and_copy.ipynb
Quora_duplicate_questions_classification.ipynb		Quora_duplicate_questions_classification.ipynb
README.md		README.md
Time_series_and_RNN.ipynb		Time_series_and_RNN.ipynb
advanced numpy and Ipython system .ipynb		advanced numpy and Ipython system .ipynb
advanced-pandas.ipynb		advanced-pandas.ipynb
ch-9 Plotting and Visualization.ipynb		ch-9 Plotting and Visualization.ipynb
cha-8 Data Wrangling.ipynb		cha-8 Data Wrangling.ipynb
commonLiterature.ipynb		commonLiterature.ipynb
duplicate-question-classification.ipynb		duplicate-question-classification.ipynb
kaggle data survey .ipynb		kaggle data survey .ipynb
pandas chp-7.ipynb		pandas chp-7.ipynb
pandas.ipynb		pandas.ipynb
recommendation_with_knn.ipynb		recommendation_with_knn.ipynb

Provide feedback