Exploring interactive machine learning model dashboards/apps that can be used to better understand and analyze the important factors in which the model works, or for visualisations.
| notebooks | description | additional related data | |
|---|---|---|---|
| 1 | MNIST_Tensorboard | Using Tensorboard | logs folder |
| 2 | app.py | Flask app | templates folder, finalized_pred_model.sav |
| 3 | Tableau | link to Tableau Public |
Getting a machine to recognise handwritten single digits is difficult. The aim here is to analyse the features of handwritten digits, using a CNN model, visualised in TensorBoard.
Clinicians need a way to assess the likelihood of a patient being readmitted due to diabetes. So the aim is to create an API framework using Flask/python, that accepts the inputs into a Logistic Regression model, and return the re-admission probability of patients with diabetes.
An audiobook company would like information on the review score, number of reviews, etc, over time. As it is good practice to first design a dashboard before working in Tableau, the first goal is to know what information would be useful for the company; Once we know the information needed, it will be easier to create/design the necessary charts.
- What are the ratings and average rating?
- Which are the audiobooks that collected the most reviews? This provides information on best-sellers.
- What is the reviews to sales ratio? This provides information on whether people who buy audiobooks also leave reviews.