DevMesh

Exhibition Art - Shipping Cost Predictor

Deepak Joshi

Deepak Joshi

New Delhi, Delhi

4 0
  • 0 Collaborators

My project is the complete analysis of the data of a Shipping Company. The main aim is to predict the cost of shipping provided the required entries were given but along with the EDA, it becomes much more than just the Cost Predictor. ...learn more

Project status: Concept

oneAPI, Artificial Intelligence

Groups
Student Developers for oneAPI, IOSC-BVP

Intel Technologies
Intel Opt ML/DL Framework, Intel Python, Other

Docs/PDFs [1]Code Samples [1]Links [1]

Overview / Usage

My project is the complete analysis of the data of a Shipping Company. The main aim is to **predict the cost of shipping **provided the required entries were given but along with the EDA, it becomes much more than just the Cost Predictor.

  1. This will help the company to increase the Retention Rate of customers.
  2. Reduce Delivery costs.
  3. The company can also target specific regions accordingly to demand (i.e Opening bigger warehouses in more demanding areas instead of random areas or near multiple normal demanding areas) and many more.
  4. Helps in reducing a few unnecessary practices.

Methodology / Approach

I used Supervised Machine Learning to predict the cost of Shipping the sculptures.

  1. Cleaning the Data using basic EDA and undestood the data types of features.
  2. Verified the Correlation between different features using Pearson Correlation and Chi-squared Test.
  3. EDA helped us understand that data was not ready to be used in the ML model so **Imputation **was done to fill missing values or N.A. values.
  4. New Features were created from already existing features to remove the extra information which might not impact our model's prediction.
  5. Few Features were removed from the Dataset before feeding it to ML models (Customer Name, Customer ID,)
  6. A lot of Features in our dataset it categorical types and a few algorithms do not work on Categorical features so they were encoded into numerical values.
  7. Few ML algorithms expect the **Scaled **data so data is scaled before putting it through the ML model.
  • ML model used:
  1. Linear Regression
  2. Decision Tree Regressor
  3. RandomForest Regressor
  4. K-Neighbour Regressor
  5. Gradient Boosting Regressor
  6. XGBoost Regressor
  7. AdaBoost Regressor
  • Also performed GridSearchCV from hyperparameter tuning.
  • **Results **in the form of error were taken and compared among ML models.
  • **Parameters **for choosing model: MAE,R2_score,MSE.

Technologies Used

  • Built Using: Python
  • Libraries Used: Numpy, Pandas, Matplotlib,Seaborn,SweetViz,Sklearn,Scipy,FeatureEngine
  • Domain: Machine Learning & EDA

Documents and Presentations

Exhibit-Art__1_.pdf

Repository

https://github.com/deepakjoshi2k/Machine-Learning-Exhibit-Art-Shipping-

Comments (0)