Introduction:

      A Machine Learning-based system to predict the weight of the baby which is going to born by using the various data points from a pregnant lady.

Purpose: Help to arrange better medical care for the child in advance.

Connection to Business: The ntelligent Machine learning model can be easily transformed into a complete software solution and can be provided to hospitals as a service with subscriptions. It can boost the quality of any hospital and save a lot of time and other things by avoiding any emergency treatment.

Achievement:   

• Achieved 94% accuracy, which is fantastic.

• Successfully deployed on Google Cloud Platform with the model pipeline for contentious training.

 

Challenges: 

1. Understanding the impact of each feature on the baby’s weight was required a good study on medical content.

2. Implementation of with-ultrasound and without-ultrasound scenario was a tricky challenge.

 

 Learning:

1. Learn the Google Transfer Appliance to handle large data.

2. ETL Pipelines patterns

 

 Contributions:

1. I have implemented the main logic of the model

2. Finalize the complete workflow to deploy the model.

 

 Data Size:  ~25GB