Modelbeoordeling

Karina beoordeelt drie verschillende Supervised Machine Learning-modellen om frauduleuze banktransacties te detecteren. Karina weet dat een zwak punt van de trainingsgegevens is dat er veel meer voorbeelden van geldige transacties zijn dan van frauduleuze transacties.

Rechts zie je hoe elk model een transactie als frauduleus of geldig voorspelde, vergeleken met hoe vaak de transactie daadwerkelijk frauduleus of geldig was.

Volgens deze gegevens: welk model is het beste en waarom?

Deze oefening maakt deel uit van de cursus

Data Science voor het bedrijfsleven

Cursus bekijken

Praktische interactieve oefening

Zet theorie om in actie met een van onze interactieve oefeningen.

Begin met trainen

Deze oefening maakt deel uit van de cursus

Data Science voor het bedrijfsleven

SkillTag.level.beginnerSkillTag.label

4.8+

Begin de cursus gratis

We'll start the course by defining what data science is. We'll cover the data science workflow, and how data science is applied to real-world business problems. We'll finish the chapter by learning about ways to structure your data team to meet your organization's needs.

Exercise 1: What is Data Science?Exercise 2: Customer Segmentation Workflow Exercise 3: Building a customer service chatbot Exercise 4: Improving OKRs Exercise 5: Applications of Data Science Exercise 6: Assigning data science projects Exercise 7: Investment research Exercise 8: Building a data science team Exercise 9: Interpreting a team sprint Exercise 10: Editing a job posting Exercise 11: Matching skills to jobs Exercise 12: Classifying data tasks

Now that we understand the data science workflow, we'll dive deeper into the first step: data collection. We'll learn about the different data sources your company can draw from, and how to store that data once it's collected.

Exercise 1: Data sources and risks Exercise 2: Classifying data for security Exercise 3: Creating web data events Exercise 4: Protecting PII Exercise 5: Solicited data Exercise 6: Identifying question purpose Exercise 7: Validating focus group feedback Exercise 8: Net Promoter Score Exercise 9: Collecting additional data Exercise 10: Sorting data sources Exercise 11: Asthma frequency Exercise 12: Data storage and retrieval Exercise 13: Cloud platforms Exercise 14: Querying a database Exercise 15: Which type of database?

In this chapter, we'll discuss ways to explore and visualize data through dashboards. We'll discuss the elements of a dashboard and how to make a directed request for a dashboard. This chapter will also cover making ad hoc data requests and A/B tests, which are a powerful analytics tool that de-risk decision-making.

Exercise 1: Dashboards Exercise 2: Classifying dashboard elements Exercise 3: Improving a dashboard Exercise 4: Choosing the right dashboard Exercise 5: Ad hoc analysis Exercise 6: Filling out an ad hoc request Exercise 7: Classifying requests Exercise 8: A/B Testing Exercise 9: Creating an A/B testing workflow Exercise 10: Sample size Exercise 11: Intermediate results

In this final chapter, we'll discuss the buzziest topic in data science: machine learning! We'll cover supervised and unsupervised machine learning, and clustering. Then, we'll move on to special topics in machine learning, including time series prediction, natural language processing, deep learning, and explainable AI!

Exercise 1: Supervised Machine Learning Exercise 2: Wanneer gebruik je Supervised Learning Exercise 3: Features en labels Exercise 4: Modelbeoordeling

Huidige oefening

Exercise 5: Clustering Exercise 6: Supervised vs. Unsupervised Exercise 7: Aantal clusters kiezen Exercise 8: Speciale onderwerpen in Machine Learning Exercise 9: Machine-learningtaken classificeren Exercise 10: Sentimentanalyse Exercise 11: Deep Learning en Explainable AI Exercise 12: De juiste oplossing vinden Exercise 13: Moet ik Deep Learning gebruiken?