Order from us for quality, customized work in due time of your choice.
(10 points)
Load the dataset (credit_scores.csv)
Delete the following features f
(10 points)
Load the dataset (credit_scores.csv)
Delete the following features from the dataset: ″Name″, ″SSN″, ″ID″, ″Customer_ID″
Set the Credit_Score feature as the target variable, and the remaining features as the input variables.
Divide the dataset by 80% training and 20% testing using the random seed number of 1.
(10 points, if you will use pipeline +10 points),
Do the following pre-processing tasks:
Fill the missing values in the numerical columns using the mean values of the respected columns and scale the numerical columns using the standard scaler.
Fill the missing values in the categorical columns using the most frequent values of the respected columns and convert the categorical columns into numerical columns using the label encoders.
(20 points, if you will use pipeline +10 points)
Develop a Support Vector Machine model to make credit score classification.
Fine-tune the following hyperparameters: kernel value of ′rbf′ or ′linear′ and Complexity value of 0.01,10,20
Report the accuracy of your best model.
Retrain your best model using the whole dataset, and save your best model as a file.
(30 points, if you will use pipeline +10 points)
Here, you can use any Python IDE.
Develop a web application to use your classification model in the production environment.
Or develop a web API and its consumer application or its CURL command to test the WEB API.
(10 points)
Upload your Web application or Web API to a host on the internet and share a working web link.
Finally, make sure you have done a public GitHub repo and you have uploaded all files there, and if you use cloud notebooks you make them public and share links to the GitHub repo and cloud notebooks from the Platon. Double-check if your links are publicly accessible or not.
Make sure after dateline of the exam don’t make any change in your GitHub and public Notebooks.
Instruction:
Go to Kaggle, register to Kaggle, follow the user ″SelcukCan″ on the Kaggle, open a new Notebook rename it to your name and surname, and make your notebook public.
Or you can use Google Colab Notebook, make sure your Colab notebook is shared publicly.
Or you can use any desktop Notebook application.
(10 points)
Load the dataset (credit_scores.csv)
Delete the following features from the dataset: ″Name″, ″SSN″, ″ID″, ″Customer_ID″
Set the Credit_Score feature as the target variable, and the remaining features as the input variables.
Divide the dataset by 80% training and 20% testing using the random seed number of 1.
(10 points, if you will use pipeline +10 points),
Do the following pre-processing tasks:
Fill the missing values in the numerical columns using the mean values of the respected columns and scale the numerical columns using the standard scaler.
Fill the missing values in the categorical columns using the most frequent values of the respected columns and convert the categorical columns into numerical columns using the label encoders.
(20 points, if you will use pipeline +10 points)
Develop a Support Vector Machine model to make credit score classification.
Fine-tune the following hyperparameters: kernel value of ′rbf′ or ′linear′ and Complexity value of 0.01,10,20
Report the accuracy of your best model.
Retrain your best model using the whole dataset, and save your best model as a file.
(30 points, if you will use pipeline +10 points)
Here, you can use any Python IDE.
Develop a web application to use your classification model in the production environment.
Or develop a web API and its consumer application or its CURL command to test the WEB API.
(10 points)
Upload your Web application or Web API to a host on the internet and share a working web link.
Finally, make sure you have done a public GitHub repo and you have uploaded all files there, and if you use cloud notebooks you make them public and share links to the GitHub repo and cloud notebooks from the Platon. Double-check if your links are publicly accessible or not.
Order from us for quality, customized work in due time of your choice.