Laten we beginnen. Het is Gratis
of registreren met je e-mailadres
ML 1 Door Mind Map: ML 1

1. Modellen

1.1. Linear regression

1.2. Logistic regression

1.3. Decision Tree

1.4. Ensemble models

1.4.1. Random Forest

1.5. Deep learning

2. Artifacts

2.1. Data

2.2. Predictive model

2.3. Code

3. Evaluatie methoden

3.1. Classificatie

3.1.1. Confusion matrix

3.1.2. Area under the rock curve

3.1.3. Recall/ TPR

3.1.4. TNR

3.1.5. Accuracy

3.2. Regressie

3.2.1. Root mean square error (RMSE)

3.2.2. Mean square error (MSE)

3.2.3. Mean absolute percentage error (MAPE)

4. Overfitting

4.1. Oorzaken

4.1.1. Teveel iteraties

4.1.2. Te flexibele hypothese

4.1.3. Te weinig data samples

4.1.4. Teveel features

4.2. Remedies

4.2.1. Feature selection

4.2.2. Meer data samples

4.2.3. Minder iteraties

4.2.4. Simpeler hypothese

4.2.5. Regularisatie

4.2.6. Hyperparameter tuning

5. Data

5.1. Categorical

5.1.1. One-hot encoding

5.2. Numerical

6. Explainability

6.1. Feature importance

6.2. Feature attribution

7. Categorie

7.1. Supervised

7.2. Unsupervised

7.3. Reinforcement learning

8. Stappen in ML

8.1. Data verzamelen

8.2. Exploratory data analysis

8.2.1. Begrip opbouwen

8.2.1.1. Business understanding

8.2.1.2. Visualisaties

8.2.2. Verificatie

8.2.2.1. Missende data

8.2.2.2. Outliers

8.2.2.3. Schoonmaken van de data

8.2.3. Correlatie zoeken

8.3. Hypothese kiezen

8.4. Parameter learning (theta)

8.4.1. Kiezen cost function

8.4.1.1. Root mean square error (RMSE)

8.4.1.2. Mean square error (MSE)

8.4.1.3. Mean absolute percentage error (MAPE)

8.4.2. Gradient descent

8.4.2.1. Stochastic

8.4.2.2. Batch

8.5. Evalueren

9. MLOps

9.1. Volledige lifecycle van het product

9.1.1. Collect

9.1.2. Curate

9.1.3. Transform

9.1.4. Validate

9.1.5. Explore

9.1.6. Train

9.1.6.1. Continuous training

9.1.7. Evaluate

9.1.8. Formulate

9.1.8.1. Business objective

9.1.8.2. Cost of mistakes

9.1.8.3. Data availability

9.1.8.4. Evaluation metrics

9.1.8.5. ML - canvas

9.2. Extensie van DevOps

9.2.1. Continuous integration

9.2.1.1. Kwaliteit- en consistentietests

9.2.2. Continuous delivery

9.2.2.1. Snel itereren

9.2.2.2. Snel feedback

10. Hyperparameters

10.1. Batch size

10.2. Number of iterations

10.3. Lambda

10.4. Learning rate

11. Correlatie

11.1. Numeriek - numeriek

11.1.1. Scatterplot

11.2. Numeriek - categorisch

11.2.1. Bar chart

11.2.1.1. Confidence interval

11.3. Categorisch - categorisch

11.3.1. Grouped bar chart

11.3.2. Chi squared