Skip to content
DeepForgeby svx
Intermediaterankingrecommendation-systemsretrievalevaluation

Ranking & Recommendation Systems

26 problems·~9h·4 stages

Search relevance and recommender systems, built from the retrieval stage to the ranking metrics. You will implement TF-IDF, inverted indexes, BM25, collaborative filtering, and NDCG from scratch. A strong fit for ML engineers working on feeds, search, or personalization.

Your progress

0/26 · 0%

Path readiness 0%Easy 0% · Medium 0% · Hard 0%

Continue path · Implicit Feedback Confidence

What you will learn

  • Build retrieval with TF-IDF, inverted indexes, and BM25
  • Score rankings with precision@k, MAP, MRR, and NDCG
  • Recommend with neighbourhood and matrix-factorization methods
  • Blend signals with rank fusion and graph similarity

Before you start

1Candidate Signals

0/7

The raw signals behind feed ranking: implicit-feedback confidence, precision at k, mean reciprocal rank, term frequency, inverse document frequency, and graph similarity scores.

Stage checkpoint

0/2 to pass

A checkpoint is a short retrieval test drawn from this stage: solve 2 of 3 from memory to mark the stage complete. The stage stays open either way.

Tried this set and did not pass? Record it — the next set swaps in fresh problems from the same stage.
LabRecommended next stepLogistic Regression, From ScratchEasy

2Retrieval

0/6

Find candidates fast: TF-IDF, inverted indexes, positional search, PMI collocations, Adamic-Adar scores, and BM25.

Stage checkpoint

0/2 to pass

A checkpoint is a short retrieval test drawn from this stage: solve 2 of 3 from memory to mark the stage complete. The stage stays open either way.

Tried this set and did not pass? Record it — the next set swaps in fresh problems from the same stage.
LabRecommended next stepSpam FilterMedium

3Ranking Metrics and Graph Ranking

0/7

Measure and rank: lift at k, mean average precision, NDCG, ROUGE-2, HITS hub and authority scores, Brandes betweenness, and PageRank to tolerance.

Stage checkpoint

0/2 to pass

A checkpoint is a short retrieval test drawn from this stage: solve 2 of 3 from memory to mark the stage complete. The stage stays open either way.

Tried this set and did not pass? Record it — the next set swaps in fresh problems from the same stage.
LabRecommended next stepLogistic Regression, From ScratchEasy

4Collaborative Filtering

0/6

Learn user and item vectors: user-based and item-based prediction, matrix-factorization SGD, association lift, feature-crossing hashes, and weighted reciprocal-rank fusion.

Stage checkpoint

0/2 to pass

A checkpoint is a short retrieval test drawn from this stage: solve 2 of 3 from memory to mark the stage complete. The stage stays open either way.

Tried this set and did not pass? Record it — the next set swaps in fresh problems from the same stage.
LabRecommended next stepLogistic Regression, From ScratchEasy

Capstone

Project

Close the path with one real artifact from the catalog.

Build a Recommender System

Carry collaborative filtering and matrix factorization from this path into a working recommender.