Now we can that we have done with this data set, you have the data analysis, statistics conclusions, and model prediction, may this helps in taking the decisions in such cases. The evaluation of the model gives us a look about the performance of the model, the results was good, if we have some edits on the model by dropping some features for example maybe we could get even much better results. The model creation and training the model and have predictions, we have to evaluate the model itself to have a better look at the predictions of the model to be sure of the results and how far could we depend on the model in taking decisions. The features in our case need to be rescaled are ( weight, discount) as rest of the features we have the same range of data. Second, we have to make scaling to avoid the outliers or fake errors in creating the model and make sure the normalization of features The suitable model algorithm technique to be used in this case is logistic regression, as we want to choose between two results, zero or one, so the perfect way to get the predictions.įirst, we have to map all the categorical data and convert it to numerical data to be suitable for the regression model, In our model we have only 4 features that are needed to be mapped(warehouse_block,Gender, Product_importance, Ship_mode), we have converted them to numbers zeroes and ones, etc. For the last part, it comes always the prediction, after we have cleaned, analyzed, and summarized the data, what will we do in the future, that is why we need to build a model to help us in predicting the result based on our data.
0 Comments
Leave a Reply. |