Vous êtes sur la page 1sur 46

DATA

SCIENCE
ANALYTICS METHODOLOGY
A general framework for data mining and
analytics

Steps in the analytics methodology


• Problem Identification
• Solution Design
• Solution Implementation
• Solution Monitoring

Best Practices in Analytics


Analytics Methodology

Problem Solution Solution Solution


Definition Design Implementation Monitoring
Analytics Methodology

Problem Solution Solution Solution


Definition Design Implementation Monitoring

The “best” solution based on theory may not be the “best” solution
in practice
Analytics Methodology

Problem Solution Solution Solution


Definition Design Implementation Monitoring

The “best” solution based on theory may not be the “best” solution
in practice

1. Lack of good data or specific variables


Analytics Methodology

Problem Solution Solution Solution


Definition Design Implementation Monitoring

The “best” solution based on theory may not be the “best” solution
in practice

1. Lack of good data or specific variables


2. Solution implementation constraints
Analytics Methodology

Problem Solution Solution Solution


Definition Design Implementation Monitoring

The “best” solution based on theory may not be the “best” solution
in practice

1. Lack of good data or specific variables


2. Solution implementation constraints
3. Other business constraints
Analytics Methodology

Problem Solution Solution Solution


Definition Design Implementation Monitoring

The “best” solution based on theory may not be the “best” solution
in practice

1. Lack of good data or specific variables


2. Solution implementation constraints
3. Other business constraints

Need to understand how solution will be implemented to finalize best


solution
Solution Design
There are multiple approaches to business problems - is an analytics
solution always the best approach?
Solution Design
There are multiple approaches to business problems - is an analytics
solution always the best approach?

 To use data to drive decisions, you need to ensure you have the right
data
Solution Design
There are multiple approaches to business problems - is an analytics
solution always the best approach?

 To use data to drive decisions, you need to ensure you have the right
data

 For example: you want to predict the potential trial rate of an ultra-
luxury brand that you want to launch, but you have so far launched
brands that are more value conscious – no data for an analytical
approach.
Solution Design
There are multiple approaches to business problems - is an analytics
solution always the best approach?

 To use data to drive decisions, you need to ensure you have the right
data

 For example: you want to predict the potential trial rate of an ultra-
luxury brand that you want to launch, but you have so far launched
brands that are more value conscious – no data for an analytical
approach.

 You could try to obtain data on competitor launches or run some


qualitative research in this case, but it is not a good idea to develop
a trial % based on a different segment of population
Solution Design
Solution design may depend on end use:

1. Reporting for insights : Use historical data to understand


performance and identify patterns in data and use as basis
for future planning

2. Forecasting : Use past data collected over time for actual


forecasts of future performance

3. Predictive Analytics : Similar to forecasting, but not just


time related

4. Optimization : Develop strategies to derive optimal use of


resources given constraints
Solution Design
Solution design may depend on end use:

1. Reporting for insights : Use historical data to understand


performance and identify patterns in data and use as basis
for future planning

2. Forecasting : Use past data collected over time for actual


forecasts of future performance

3. Predictive Analytics : Similar to forecasting, but not just


time related

4. Optimization : Develop strategies to derive optimal use of


resources given constraints
Solution Design
Solution design may depend on end use:

1. Reporting for insights : Use historical data to understand


performance and identify patterns in data and use as basis
for future planning

2. Forecasting : Use past data collected over time for actual


forecasts of future performance

3. Predictive Analytics : Similar to forecasting, but not just


time related

4. Optimization : Develop strategies to derive optimal use of


resources given constraints
Solution Design
Solution design may depend on end use:

1. Reporting for insights : Use historical data to understand


performance and identify patterns in data and use as basis
for future planning

2. Forecasting : Use past data collected over time for actual


forecasts of future performance

3. Predictive Analytics : Similar to forecasting, but not just


time related

4. Optimization : Develop strategies to derive optimal use of


resources given constraints
Solution Design
Solution design may depend on end use:

1. Reporting for insights : Use historical data to understand


performance and identify patterns in data and use as basis
for future planning

2. Forecasting : Use past data collected over time for actual


forecasts of future performance

3. Predictive Analytics : Similar to forecasting, but not just


time related

4. Optimization : Develop strategies to derive optimal use of


resources given constraints
Solution Design
Solution design may depend on end use:

1. Reporting for insights : Use historical data to understand


performance and identify patterns in data and use as basis
for future planning

2. Forecasting : Use past data collected over time for actual


forecasts of future performance

3. Predictive Analytics : Similar to forecasting, but not just


time related

4. Optimization : Develop strategies to derive optimal use of


resources given constraints
Solution Design
Solution design may depend on end use:

1. Reporting for insights : Use historical data to understand


performance and identify patterns in data and use as basis
for future planning

2. Forecasting : Use past data collected over time for actual


forecasts of future performance

3. Predictive Analytics : Similar to forecasting, but not just


time related

4. Optimization : Develop strategies to derive optimal use of


resources given constraints
Solution Design
Solution design may depend on end use:

1. Reporting for insights : Use historical data to understand


performance and identify patterns in data and use as basis
for future planning

2. Forecasting : Use past data collected over time for actual


forecasts of future performance

3. Predictive Analytics : Similar to forecasting, but not just


time related

4. Optimization : Develop strategies to derive optimal use of


resources given constraints
Solution Design
Solution design may depend on end use:

1. Reporting for insights : Use historical data to understand


performance and identify patterns in data and use as basis
for future planning

2. Forecasting : Use past data collected over time for actual


forecasts of future performance

3. Predictive Analytics : Similar to forecasting, but not just


time related

4. Optimization : Develop strategies to derive optimal use of


resources given constraints
ANALYTICS METHODOLOGY
A general framework for data mining and
analytics

Steps in the analytics methodology


• Problem Identification
• Solution Design
• Solution Implementation
• Solution Monitoring

Best Practices in Analytics


Analytics Methodology

Problem Solution Solution Solution


Definition Design Implementation Monitoring

Solution Implementation
Analytics Methodology

Problem Solution Solution Solution


Definition Design Implementation Monitoring

Solution Implementation

1. Typically is the longest phase


Analytics Methodology

Problem Solution Solution Solution


Definition Design Implementation Monitoring

Solution Implementation

1. Typically is the longest phase

2. Starts with data collection


Analytics Methodology

Problem Solution Solution Solution


Definition Design Implementation Monitoring

Solution Implementation

1. Typically is the longest phase

2. Starts with data collection

3. Data Preparation and Exploratory Data Analysis will typically


form a significant part of total project effort and time
Analytics Methodology

Problem Solution Solution Solution


Definition Design Implementation Monitoring

Solution Implementation

4. Initial modeling results and calibration


Analytics Methodology

Problem Solution Solution Solution


Definition Design Implementation Monitoring

Solution Implementation

4. Initial modeling results and calibration

5. Finalization of models post calibrations


Analytics Methodology

Problem Solution Solution Solution


Definition Design Implementation Monitoring

Solution Implementation

4. Initial modeling results and calibration

5. Finalization of models post calibrations

6. Interpretation and framing of results in business context


Solution Implementation
Once the best solution approach has been identified, the
actual solution building and implementation phase starts

This phase includes:

1. all the data exploration and preparation phases,


2. the analytical techniques application, and
3. the final implementation of the derived solution

This phase typically accounts for 60 – 80% of total project time


Solution Implementation
Typical Steps within Solution Implementation

1. Data Exploration
2. Data Preparation
3. Data Partitioning
4. Model Building – Test Data Set
5. Initial Model
6. Model Iterations and Tweaking
7. Final Model
8. Validation Model
9. More Iterations If required based on Validation Model Results
10. Interpretation of results into business friendly context
11. Solution implementation
ANALYTICS METHODOLOGY
A general framework for data mining and
analytics

Steps in the analytics methodology


• Problem Identification
• Solution Design
• Solution Implementation
• Solution Monitoring

Best Practices in Analytics


Analytics Methodology

Problem Solution Solution Solution


Definition Design Implementation Monitoring

Solution Monitoring
Analytics Methodology

Problem Solution Solution Solution


Definition Design Implementation Monitoring

Solution Monitoring

1. Very important phase that is sometimes ignored


Analytics Methodology

Problem Solution Solution Solution


Definition Design Implementation Monitoring

Solution Monitoring

1. Very important phase that is sometimes ignored

2. Once solution is implemented, need to ensure that original


assumptions / constraints still hold
Analytics Methodology

Problem Solution Solution Solution


Definition Design Implementation Monitoring

Solution Monitoring

1. Very important phase that is sometimes ignored

2. Once solution is implemented, need to ensure that original


assumptions / constraints still hold

3. Constant monitoring of solution to ensure changes are


accurately captured and assessed
Analytics Methodology

Problem Solution Solution Solution


Definition Design Implementation Monitoring

Solution Monitoring

1. Very important phase that is sometimes ignored

2. Once solution is implemented, need to ensure that original


assumptions / constraints still hold

3. Constant monitoring of solution to ensure changes are


accurately captured and assessed

4. Solutions age – need to track reliability over time


Solution Monitoring
Solution monitoring is required to ensure that solution
supplied is continuously providing expected results

For example:

We have devised a model to predict probability of default -


if the characteristics of the underlying applicant pool changes, or there
are other external factors affecting population behavior,

you will need to modify the solution or potentially come up with a


new one
ANALYTICS METHODOLOGY
A general framework for data mining and
analytics

Steps in the analytics methodology


• Problem Identification
• Solution Design
• Solution Implementation
• Solution Monitoring

Best Practices in Analytics


Analytics Best Practices
1. Models should be as simple as possible
• Example : Impact of Age on Sales
• Ages 20-30 v/s Ages 50-60
Analytics Best Practices
1. Models should be as simple as possible
Analytics Best Practices
1. Models should be as simple as possible

2. Model results and recommendations need to be actionable


Analytics Best Practices
1. Models should be as simple as possible

2. Model results and recommendations need to be actionable

3. Models need to be rigorously validated on multiple sample sets


if possible
Analytics Best Practices
1. Models should be as simple as possible

2. Model results and recommendations need to be actionable

3. Models need to be rigorously validated on multiple sample sets


if possible

4. The most critical three things to remember:


Analytics Best Practices
1. Models should be as simple as possible

2. Model results and recommendations need to be actionable

3. Models need to be rigorously validated on multiple sample sets


if possible

4. The most critical three things to remember:

– Business Knowledge, Business Knowledge,


Business Knowledge!
THANK YOU

Vous aimerez peut-être aussi