Recording: https://digitalmarketinginstitute.in/webinar-05-marketing-analytics-with-python-jupyter-notebook/
Website: https://digitalmarketinginstitute.in/
In this webinar, Mr Abhishek Kumar (Ex. Sr. Consultant, Adobe) is sharing his expertise around data analysis using Python and Jupyter Notebook. Abhishek brings along over 10 years of experience in consulting.
Abhishek discussed key-concepts in predictive analytics, how organizations use Python and Jupyter notebook for data analysis.
Python and Machine Learning
Segmentation
Association Rules
Python & Jupyter Notebook
Question & Answer
3. Agenda
- Recap from the last week
- Segmentation
- What is segmentation
- Why do we need segmentation?
- How segmentation is done ?
- Association rules
- Apriori Algorithm
- Why do we need Association rules
- Digital Marketing Concepts
4. Recap from last week
- Web Analytics concepts
- Tools for web analytics and data analysis
- Jupyter Notebook & Python
- Data Cleaning
- Exploratory data analysis
- Linear regression model
5. Data Segmentation
● What is segmentation?
● Why do we need segmentation?
● How do we do segmentation?
○ K-means clustering approach
6. Association Rules
● What is association rules?
● Why do we care about association rules?
○ Discover interesting relation between variables in dataset
● Algorithm for association rules
○ Apriori algorithm
7. Association Rules
{eggs} ---> {bread} 1000
A = Bread purchases = 500 transactions
C = Egg Purchases = 350 transaction
A-->C = Both bread and egg purchased in 150
1. Support
2. Confidence
3. Lift
8. Support
It is a relative frequency of an item in a dataset.
Support of an item = transaction containing that item/total number of
transaction
9. Confidence
Confidence is probability of seeing the consequent item within data given that
the data also contains the antecedent item.
In other words-
How likely it is for 1 item to be purchased given that another item is purchased
Confidence(A-->C) = Support(A→C)/suport(A)
10. Lift
It measures how much more often the antecedent and consequent occur
together rather than them occurring independently
lift(A-->C) = confidence(A--->C)/support(C)
11. Digital Marketing
- Workflow of digital marketing
- Tools used in digital marketing
- Social media platform
- Email marketing tool
- Analytics tool
- Content marketing tool and design tool
12. Next Steps for you!
- Revise the session and pick your favorite topic such as Web analytics,
Linear regression, python skills etc.
- Start working on that topic and write us back if you have any questions
Abhishek Kumar- abhikum@bu.edu