Scribd
Upload
58 views5 pages

Atoti Python

This document discusses using machine learning and business intelligence (BI) analytics with a car insurance dataset to understand price elasticity and optimize pricing. It describes building predictive models to forecast sales based on price and other factors. Simulation scenarios are created in the atoti BI platform to analyze how adjusted pricing impacts key performance indicators and business metrics. The goal is to bridge the gap between machine learning and business outcomes.

Uploaded by

bieito1968
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
58 views5 pages

Atoti Python

This document discusses using machine learning and business intelligence (BI) analytics with a car insurance dataset to understand price elasticity and optimize pricing. It describes building predictive models to forecast sales based on price and other factors. Simulation scenarios are created in the atoti BI platform to analyze how adjusted pricing impacts key performance indicators and business metrics. The goal is to bridge the gap between machine learning and business outcomes.

Uploaded by

bieito1968
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 5

What-if simulation in

Python
Integrate machine learning with BI
analytics to realize business values
for car insurance
Using data from https://www.kaggle.com/ranja7/vehicle-
insurance-customer-data
Different products have different
sensitivities to price changes.

Problem statement:

What would be
the lowest price
that would bring
in the maximum Quantity
realized revenue?

How does a
price fluctuation
Price affect demand?

Unravel price elasticity of demand

Traditional Price
Technical Pricing Fixed profit margin
(Bare minimum price
to bear the potential
claim) Pricing Optimization
Optimized Pricing
to maximize profits

Price Elasticity
Model

Shared by TRS @atoti


Machine Learning

Predictive Modeling Feature selection


• Random forest classifier • ['Price', 0.20924321]
• KNeighbors Classifier • ['Driver Age', 0.06532796]
• Gaussian Naive Bayes Classifier • ['Employed', 0.061190005]
• Support Vector Machine based Classifier • ['Personal Auto', 0.056500446]
• XG Boost Classifier • ['Suburban', 0.04785541]
• [‘Special Auto', 0.0455167]
• ['Premium', 0.036693245]

Sales prediction

Load sales prediction into

Adjust pricing using Pareto principle


Perform BI analytics • Price ↑ 20% for policies with sales
probability > 80%
• Price ↓ 20% for policies with sales
probability < 20%

Price adjustment by feature


performance

Analyze impact across scenarios Repeat price adjustment according to


Pareto principle, run prediction and load
results into atoti as a simulation scenario.
Load prediction for adjusted
Compare KPIs for ML models post price
pricing as a new scenario in atoti
adjustments without having to redefine
formulae.

Shared by TRS
@atoti
Dashboarding with
atoti is a free Python library that allows users to create
a collaborative BI analytics platform.
It comes with a web application for dashboarding.

Analyze business metrics of initial data by various features

Interactively compare across scenarios using filters

@atoti
Use to bridge the gap
between ML and business KPIs

See the notebook at: Learn more about atoti at:

https://github.com/atoti/
notebooks/tree/master/ https://www.atoti.io/
notebooks/price-elasticity

Follow atoti on:

LinkedIn Youtube

https://www.linkedin.com/company/atoti https://www.youtube.com/c/atoti

Twitter Medium

https://twitter.com/atoti_io https://medium.com/atoti

Reach out to atoti on:

GitHub Discussions

https://github.com/atoti/atoti/discussions

@atoti

You might also like