Most Advanced AI

We can generate Machine Learning programs from English descriptions. It can do pre-processing of data such as data transformation, scaling, and dealing with missing value. It can split data based on the user’s requirement. It can generate programs for both classification and regression. It can fit various models with different parameters. It can hyper-optimize various parameters of the selected model. To further improve the accuracy of the model, it can ensemble multiple models. The model can be saved to the hard disk using pickle. Using this model, we can predict the unknown target variable.

Below we give four examples:

Example 1: Classification - Training
file liver.csv, columns = [Age,Gender,Total Bilirubin,Direct Bilirubin,Alkaline Phosphotase,SGPT,SGOT,Total Protiens,Albumin,A/G Ratio,Target]

Train to find whether the person is liver patient or not based on all columns except Gender and Target using catboost. Use 200 iterations.
Display accuracy score and classification report.
Save the model in 'liver_catboost.pkl' using pickle.

Example 2: Classification - Prediction
file liver_test.csv, columns = [Age,Gender,Total Bilirubin,Direct Bilirubin,Alkaline Phosphotase,SGPT,SGOT,Total Protiens,Albumin,A/G Ratio]

Read the model 'liver_catboost.pkl' using pickle.
Predict whether the person is liver patient or not based on all columns except Gender using model 'liver_catboost.pkl'. Save it to "liver_test_result.csv".

Example 3: Regression - Training
file diamond_training.csv, columns = [carat, cut, color, clarity, depth, table, x, y, z, price]

Train to compute the price of diamond based on all columns except 'price' using catboost regressor.
Target column is 'price'. Category columns are 1,2,3. Use 1500 iterations. Use loss function 'rmse'.
Display accuracy on training data using rmse, mae, r2.
Save the model in 'diamond_catboost.pkl' using pickle.

Example 4: Regression - Prediction
file diamond_test.csv, columns = [carat, cut, color, clarity, depth, table, x, y, z]

Read the model 'diamond_catbook.pkl' using pickle.
Predict the price of the diamonds from carat, cut, color, clarity, depth, table, x, y and z using model 'diamond_catbook.pkl'. Save it to "diamond_test_result.csv".

You have not logged in.
Please login by clicking on the link: Login