This project uses Machine learning and Information Retrival techniques to detect diseases based on symptoms and provide more details about the top fetched diseases including treatment recommendation.
-
TF-IDF
● We have the following files for TF-idf interaction Application.
○ combination (3).csv
○ cosine_results.py
○ generatesymptoms.py
○ Interaction_TF_IDF_Cosine.ipynb
○ normal (3).csv
○ synonyms.py
○ tf_idf_result.py
○ tokenizer.py
● Keep all the files in the same folder as Interaction_TF_IDF_Cosine.ipynb and then run the lines in this jupyter file in a sequential manner.
● If you want to run on the colab then upload the Interaction_TF_IDF_Cosine.ipynb and upload all the above files in the same runtime cycle.\ -
ML interaction
● We have the following files for ML interaction Application.
○ combinational (3).csv
○ decisionT.py
○ IRadaboost.py
○ IRgdb.py
○ IRsvm.py
○ IRsvm.py
○ IRxgb.py
○ knneigbh.py
○ LR.py
○ mnb.py
○ multiLP.py
○ normal (3).csv
○ randomrf.py
○ synonyms.py
○ IR_final_ML_interaction_.ipynb
● Keep all the files in the same folder as IR_final_ML_interaction_.ipynb and then run the lines in this jupyter file in a sequential manner.
● If you want to run on the colab then upload the IR_final_ML_interaction_.ipynb and upload all the above files in the same runtime cycle.\
Models | Accuracy |
---|---|
Multi Layer Perceptron | 89.42 % |
Decision Tree | 72.62 % |
Random Forest | 89.97 % |
Logistic Regression | 89.88 % |
K nearest Neighbour | 89.97 % |
Support Vector Machine | 87.99 % |
Multinomial Naive Bayes | 81.30 % |
Gradient Boosting Machine | 80.94 % |
Extreme Gradient Boosting Machine | 78.83 % |