Machine learning

Machine learning mind map

Get Started. It's Free
or sign up with your email address
Machine learning by Mind Map: Machine learning

1. Backpropagation

2. Activation functions

2.1. ReLu

2.2. Sigmoid/logistic

2.3. Binary

2.4. Tanh

2.5. Softmax

3. Regularization

3.1. L1 norm/Manhattan distance

3.2. L2 norm/Euclidean distance

3.3. Early stopping

3.4. Dropout

3.5. Data augmentation

4. Data processing (https://github.com/dformoso/machine-learning-mindmap)

4.1. Data types

4.2. Data exploration

4.3. Feature cleaning

4.4. Feature imputing

4.4.1. Hot-deck

4.4.2. Cold-deck

4.4.3. Mean substitution

4.4.4. Regression

4.5. Feature engineering

4.5.1. Decompose

4.5.2. Crossing

4.6. Feature selection

4.6.1. Correlation

4.6.2. Dimensionality reduction

4.6.3. Importance

4.7. Feature encoding

4.8. Feature normalization/scaling

4.9. Dataset construction

4.9.1. Training dataset

4.9.2. Test dataset

4.9.3. Validation dataset

5. Neural network

5.1. Types

5.1.1. CNN

5.1.1.1. Visualize layers

5.1.2. RNN

5.1.2.1. recurrent units

5.1.2.1.1. GRU

5.1.2.1.2. LSTM

5.1.2.1.3. Nested LSTM

5.1.3. NLP

5.1.3.1. Attention

5.1.3.2. Term-document matrix

5.1.3.3. topic modelling

5.1.3.3.1. matrix decomposition techniques

5.1.3.4. Matrix factorizations

5.1.3.4.1. Singular Value Decomposition (SVD)

5.1.3.4.2. Non-negative Matrix Factorization (NMF)

5.1.3.5. Pre-processing

5.1.3.5.1. Removing words

5.1.3.5.2. getting root of word

5.1.3.5.3. Normalization of term counts

5.2. Speech recognition

5.2.1. Connectionist Temporal Classification

5.2.1.1. Align text to audio

5.2.1.2. Align text to audio

6. Monte carlo tree search

7. Methods

7.1. tree-based

7.1.1. Xgboost

7.2. Regression

8. Gradient descent

8.1. Stochastic gradient descent

8.1.1. SGD with restarts

8.2. GD with Momentum

9. Performance

9.1. Loss

9.1.1. Cross-Entropy

9.1.2. Logistic

9.1.3. Quadratic

9.1.4. D1-loss

9.2. metrics

9.2.1. classification

9.2.1.1. Accuracy

9.2.1.2. Precision

9.2.1.2.1. Precision= True_Positive/ (True_Positive+ False_Positive)

9.2.1.3. Recall

9.2.1.3.1. Recall= True_Positive/ (True_Positive+ False_Negative)

9.2.1.3.2. Sensitivity

9.2.1.4. True negative rate

9.2.1.4.1. TN/(TN+FP)

9.2.1.4.2. Specificity

9.2.1.5. F1-score

9.2.1.5.1. Combines precision and recall

9.2.1.6. ROC

9.2.1.6.1. binary classifier

9.2.1.6.2. True positive rate vs false positive rate

9.2.1.6.3. for various thresholds

9.2.1.7. AUC

9.2.1.7.1. calculates area under ROC curve

9.2.1.7.2. btw. 0 and 1

9.2.2. regression

9.2.2.1. Mean squared error

9.2.2.2. Mean absolute error

9.2.3. ranking

9.2.3.1. MRR

9.2.3.2. DCG

9.2.3.3. NDCG

9.2.4. statistical

9.2.4.1. Correlation

9.2.5. computer vision

9.2.6. nlp

9.2.7. deep learning related

9.2.7.1. Inception score

9.2.7.2. Frechet Inception distance

10. Statistics

10.1. Bayesian statistics

10.2. Hypothesis testing

10.2.1. Significance level

10.2.1.1. Significance level determines the level that we want to believe in the null hypothesis

10.2.2. P-value

10.2.2.1. The probability of the observed value is called p value

10.2.3. Population, sample, estimator

10.2.4. Probability density distribution

10.2.4.1. The probability density distribution (PDF) is used to specify the probability of the random variable falling within a particular range of values

10.2.5. Central limit theorem

10.2.5.1. Central Limit Theorem states that when the sample size is large, the sample mean of the independent random variable follows normal distribution

10.3. Statistical tests

10.3.1. Difference?

10.3.1.1. comparisons of

10.3.1.1.1. means

10.3.1.1.2. variance

10.3.2. Relationship?

10.3.2.1. Independent variable?

10.3.2.1.1. yes

10.3.2.1.2. no

10.3.3. categorical

10.3.3.1. Chi-Square

10.3.3.1.1. Chi-square tests check if distributions of categorical variables differ from each other

10.4. Gaussian process

10.4.1. gives

10.4.1.1. uncertainty estimates

10.5. Descriptive statistics

10.5.1. central tendency

10.5.1.1. mean

10.5.1.2. median

10.5.2. Spread

10.5.2.1. standard deviation

10.5.3. Percentiles

10.5.4. Skewness

10.5.5. Covariance & Correlation

10.6. parametric assumptions

10.6.1. 1. independent, unbiased samples

10.6.2. 2. normally distributed

10.6.3. 3. equal variance

11. Hyperparameter tuning

11.1. Bayesian optimization

12. Interpretability

12.1. Lime

12.2. Shap

12.3. Partial dependency plot

12.4. Tree interpreter