Filmregels extraheren

Nu je een beter idee hebt welke drempels je moet gebruiken in de functie apriori(), ben je klaar om associatieregels te extraheren uit de transactionele gegevensset movie_subset. Door de set filmregels te extraheren, kunnen we gebruikers aanbevelingen doen voor films. Stel dat je al een aantal films hebt gezien: welke film kun je daarna kijken? Om dit te beantwoorden, kunnen we gebruikmaken van de bundel films die door alle andere gebruikers is bekeken en hiervoor Market Basket Analysis toepassen. Bovendien kunnen sommige filmregels redundant zijn. Daarom moet je de set niet-redundante filmregels extraheren om waardevolle informatie en inzichten te krijgen over de films die door gebruikers worden bekeken.

Deze oefening maakt deel uit van de cursus

Market Basket-analyse in R

Cursus bekijken

Praktische interactieve oefening

Probeer deze oefening eens door deze voorbeeldcode in te vullen.

# Extract rules with the apriori
rules_movies = apriori(movie_trx,
                          parameter = list(___ = ___,
                                           conf = ___,
                                           minlen = 2, 
                                           target = "___"))

# Summary of extracted rules
___(___)

Code bewerken en uitvoeren

Deze oefening maakt deel uit van de cursus

Market Basket-analyse in R

SkillTag.level.intermediateSkillTag.label

4.8+

Begin de cursus gratis

What’s in your basket? In this first chapter, you’ll learn how market basket analysis (MBA) can be used to look into baskets and dig into itemsets to better understand customers and predict their needs. Using tidyverse and dplyr you’ll discover how many baskets can be created from a given set of items, and learn the power of using market basket analysis for online shopping, offline shopping, and use cases beyond retail.

Exercise 1: Market basket introduction Exercise 2: Baskets and items Exercise 3: Single basket Exercise 4: What's in the basket?Exercise 5: Item combinations Exercise 6: Number of possible baskets Exercise 7: Subsets and supersets Exercise 8: Plot number of possible baskets Exercise 9: What is market basket analysis ?Exercise 10: Two baskets Exercise 11: Multiple baskets Exercise 12: Looking at specific items

In this chapter, you’ll convert transactional datasets to a basket format, ready for analysis using the Apriori algorithm. You’ll then be introduced to the three main metrics for market basket analysis: support, confidence, and lift, before getting hands-on with the Apriori algorithm to extract rules from a transactional dataset. Lastly, You explore how the arules package is used for market basket analysis to retrieve basket rules and to help you find the most informative and relevant rules.

Exercise 1: Transactional data Exercise 2: Transactionalizing the online data frame Exercise 3: Inspector retail Exercise 4: Density of the item matrix?Exercise 5: Metrics in market basket analysis Exercise 6: The support measure Exercise 7: Confidence and lift measures Exercise 8: Changing the appearance of rules Exercise 9: The apriori algorithm Exercise 10: Apriori principle Exercise 11: Let's go shopping for rules Exercise 12: Parameters of the apriori Exercise 13: “If this then that” with the apriori Exercise 14: Playing with the appearance Exercise 15: Redundant rules Exercise 16: Interpretation of rules

Let’s get visual. In this chapter, you’ll visually inspect the set of rules you have previously extracted. Visualizations in market basket analysis are vital given that often you are dealing with large sets of extracted rules. You’ll use the arulesViz package to create barplots, scatterplots, and graphs to visualize your sets of inferred rules. You’ll then turn sets of plots into interactive plots, making it is easier to drill into the mined association rules.

Exercise 1: Let's see what's in the basket Exercise 2: What's in the basket?Exercise 3: What's most popular ?Exercise 4: Getting fancy with the visuals Exercise 5: Visualizing metrics Exercise 6: Scatterplots Exercise 7: ArulesViz plots Exercise 8: Bringing rules to life Exercise 9: From rules to graph based visualizations Exercise 10: Playing with graphs Exercise 11: Understanding the graph Exercise 12: Portability of your graph Exercise 13: Alternative rule plots Exercise 14: Group matrix based visualizations Exercise 15: Parallel coordinates plots Exercise 16: Mastering the RuleExplorer

We’re going to the movies. In this final chapter, you’ll apply everything you’ve learned as you work with a movie dataset. Using market basket analysis you’ll turn this dataset into a movie recommendation system, using information from movie transactions to understand and predict what your audience might want to watch next.

Exercise 1: Terugblik op transacties Exercise 2: Kennismaken met de MovieLens-gegevensset Exercise 3: Filmtransacties Exercise 4: Frequentieplots Exercise 5: Associatieregels ontginnen Exercise 6: Populariteit van films Exercise 7: De juiste filmparameters kiezen Exercise 8: Filmregels extraheren

Huidige oefening

Exercise 9: Transacties en regels visualiseren Exercise 10: Waar is Pulp Fiction?Exercise 11: Regels voor films visualiseren Exercise 12: Onze favoriete films als graaf Exercise 13: Het maximale halen uit marketbasketanalyse Exercise 14: Wat beïnvloedde Pulp Fiction?Exercise 15: Wat heeft Pulp Fiction beïnvloed?Exercise 16: Zet je market-basket-skills in!