Data Science in CPG industry

 http://www.ondatascience.com/datacamp-and-microsoft-data-science-certifications-a-comparison/

DataCamp and Microsoft Data Science Certifications | On Data Science

 

The below table is a summary of the topics, libraries and tools in the Data Science courses mapped to the a CPG business (my own industry):

Consumer Engagement	What you learn
•Social Media Sentiment analysis	Polarity function, qdap’s sentiment function, subjectivity lexicons in tidytext; Case study: AirBnB reviews
•Customer Service Automation	NLP word tokenization; Topic identification: bag-of-words NLTK; Case study: Building a “fake news” clasifier
•A/B Testing	A/B test design, Power, Sensitivity, Confidence intervals, Plotting difference distribution
•ROI Advertising Measurement	XGBoost, Ensemble regression, Boosted Decision trees; Kmeans clustering
•Initiative Activation	Timeseries in Pandas, Time series resampling, Rolling windows, Time series correlations
•Precision Marketing	Hierarchical clustering, t-SNE, tree visualization, PC Analysis. Case study: Build a NMF recommender systems
•Personalized consumer engagement	Kmeans, unsupervised learning,t-SNE visualization in 2D, feature transformation
•AI Powered Diagnostic and Recommendations	Hierarchical clustering, t-SNE, tree visualization, PC Analysis. Case study: Build a NMF recommender systems
Customer/Retailer Engagement
•Individualized store assortments	Kmeans, Hierarchical clustering, time series, rolling windows
•Trade Promotion Optimization	Regression, XGBoost, Decision trees, A/B test design
•Demand Forecasting	Timeseries in Pandas, Time series resampling, Rolling windows, Time series correlations, XGBoost, Regression
•Sales Forecast	Timeseries in Pandas, Time series resampling, Rolling windows, Time series correlations, XGBoost, Regression
•Shelf Share Analytics	Specify-Compile-Fit in Keras; Keras models optimization, Convolutional neural networks for image classification
•Out of Stock from POS Time series	Timeseries in Pandas, Time series resampling, Rolling windows, Time series correlations, XGBoost, Regression
•Slow moving SKUs	Timeseries in Pandas, Time series resampling, Rolling windows, Time series correlations, XGBoost, Regression
Product Supply
•Adaptive manufacturing, predictive maintenance, demand-driven production	Regression, XGBoost, Decision trees, A/B test design, Time series rolling window, Time series correlation
•Automated quality control	Regression, XGBoost, Decision trees, A/B test design
•Transportation Optimization	Degree centrality, shortest path algorithm, NetworkX library. Case study: Github connections reccomender
R&D
•Testing acceleration via silicon models	Classification with k-nearest, Regression supervised learning
Transactions optimization
•Character and Image recognition (A/P, A/R)	Specify-Compile-Fit in Keras; Keras models optimization, Convolutional neural networks for image classification
•Automated processing via similarity	Supervized machine learning , k-nearest neighbor for classification, k-fold for regression, scaling features
•Chatbot for How-to and Transactional work	NLU foundations, spaCy, scikit-learn, rasa NLU, statefulness. Case study: virtual assistant for travel planning.