new hashtag list

 

Machine Learning Algorithms

Machine Learning

ML Types
1] Supervised ML
2] Unsupervised ML
3] Reinforcement ML


List of Machine Learning Algorithm
a] Regression Algorithm
  1. Linear Regression
  2. Polynomial Regression
  3. Poisson Regression
  4. Ordinary Least Squares (OLS) Regression
  5. Ordinal Regression/Ranking Learning
  6. Support Vector Regression
  7. Gradient Descent Regression
  8. Stepwise Regression
  9. Ridge Regression (L2)
  10. Elastic Net Regression
  11. Bayesian Linear Regression
  12. Least-Angled Regression (LARS)
  13. Neural Network Regression
  14. Locally Estimated Scatterplot Smoothing (LOESS)
  15. Multivariate Adaptive Regression Splines (MARS)
  16. Locally Weighted Regression (LWL)
  17. Quantile Regression
  18. Principal Component Regression (PCR)
  19. Partial Least Squares Regression
  20. Lasso Regression (Least absoulute selection and shrinkage
    operator)





b] Classification Algorithm
  1. Decision Tree
  2. Decision Stump
  3. Naive Bayes
  4. Gaussian Naive Bayes
  5. Bernoulli Naive Bayes
  6. Multinomial Naive Byes
  7. K Nearest Neighbours(KNN)
  8. Support Vector Machine(SVM)
  9. Linear Support Vector Classifier(SVC)
  10. NuSVC
  11. Stochastic Gradient Decent [SGD]Classifier
  12. Bayesian Network
  13. Logistic Regression
  14. Zero Rule(ZeroR)
  15. One Rule [OneR]
  16. Linear Discriminant Analysis [LDA]
  17. Quadratic Discriminant Analysis [QDA]
  18. Fisher’s Linear Discriminant

c] Clustering Algorithm
  1. K-Means Clustring
  2. K- Median Clustring
  3. Mean Shift Clustring
  4. K-Modes Clustring
  5. Fuzzy K-Modes
  6. Fuzzy C-Modes
  7. Mini batch K-Means Clustring
  8. Hierarchical Clustring
  9. Expectation Maximization
  10. DBSCAN
  11. Minimum Spanning Trees
  12. Quality Threashold
  13. Gaussian Mixture Model(GMM)
  14. Spectral Clustering

d] Neural Network
  1. Perceptron
  2. Multilayer Perceptron(MLP)
  3. Recurrent Neural Network(RNN)
  4. Convolution Neural Network(CNN)
  5. Deep Belief Network(DBN)
  6. Hopfield Networks
  7. Learning Vector Quantization(LVQ)
  8. Stacked Autoencoder
  9. Boltzmann Machine
  10. Restricted Boltzmann Machine(RBM)
  11. Generative Adversarial Network(GANs)












 

Difference between Matplotllib and Seaborn Libraries of Python

Difference between matplotllib and seaborn:-
1. Functionality:
Matplotlib: Matplotlib is mainly deployed for basic plotting.
Visualization using Matplotlib generally consists of bars, pies, lines, scatter plot.
Seaborn: Seaborn will provide a variety of visualization patterns.
It will use less syntax and has easily interesting default themes.
It is used if one has to summarize data in visualizations and also show the distribution in the data.
2. Handling Multiple Figures:
Matplotlib: Matplotlib has multiple figures and can be opened but need to be closed explicitly.
plt.close() only closes the current figure. plt.close('all') would close all.
Seaborn: The Seaborn automates the creation of multiple figures.
This leads to OOM (out of memory) issues.
3. Visualization:
Matplotlib: Matplotlib is a graphics package for data visualization of Python.
It is well integrated with NumPy and also Pandas.
The pyplot module mirrors the MATLAB plotting command very closely.
Thus MATLAB users can easily transit to plotting with Python.
Seaborn: Seaborn is more integrated for working with Pandas a data frame.
It will extend the Matplotlib library for creating beautiful graphics with Python.
4. Data frames and Arrays
Matplotlib: Matplotlib will works with data frames and arrays.
It has stateful APIs for plotting the figures and aces.
They are represented by the object and therefore plot () like calls without parameters suffices and without having to manage parameters.
Seaborn: Seaborn will work with the dataset as a whole and is much intuitive than Matplotlib.
For Seaborn the replot() is the entry API with parameter to specify the type of plot which could be line, bar, or many of other types.
Seaborn is not stateful and plot() would require passing the object.
5. Flexibility:
Matplotlib: Matplotlib is customizable and powerful.
Seaborn: Seaborn will avoid a ton of boilerplate by providing default theme which are commonly used.
6. Use Cases:
Matplotlib: Pandas will use Matplotlib.
It is also neat wrapper around Matplotlib.
Seaborn: Seaborn is used for more specific use cases and also it is Matplotlib under the hood.
Also it is specially meant for statistical plotting.

##########################################################################


PLOTTING TIP

For categorical variables utilize Bar Charts* and Boxplots.
For continuous variables utilize Histograms, Scatterplots, Line graphs, and Boxplots.