Machine Learning with Python Case Study
summary
Recommendation system for holiday accommodation built with Python 3.5.1 and Jupyter using Pandas, NumPy/SciPy and scikit‑learn, trained on ~1 million bookings with extreme sparsity and weekly seasonality windows, uses implicit ALS matrix factorization over booking signals, evaluates by top-5 hit rate and cannot randomize time series data, introduces a seasonality via confidence weighting on observations and a “super‑user” encoding for no explicit user tracking, runs many iterations so C‑backed numeric libraries are required, and yields about +5–6% top‑5 improvement over the human baseline.
AI-written summary · about
chapters (8)
- 0:00Intro
- 1:52Challenges
- 4:04Approach
- 5:49Advantages Disadvantages
- 6:30Key Aspect
- 11:17Performance
- 12:16Global Representation
- 12:57Conclusion
related talks
35mCustomer Segmentation in Python
24mPredicting Hotel Cancellations with Machine Learning
01h 38mBuilding Recommender Systems Using Python
40mIntroduction to Machine Learning
44mMachine learning @ booking.com
26mLarge-Scale Recommendation System with Python and Spark
33mMachine Learning explained to humans
27mFrom Data to Insights in Minutes: Accelerating Predictive Modelling (in Python) with AutoML