#Kalman Filter #Particle Filter #Graphical Models

Advanced Machine Learning

1 Libraries 2 Graphical Models 2.0.1 D-seperation 2.1 1)Show that multiple runs of the hill-climbing algorithm can return non-equivalent Bayesian network (BN) structures. Explain why this happens. Use the Asia dataset which is included in the bnlearn package. To load the data, run data(“asia”). 2.2 2) Learn a BN from 80 percent of the Asia dataset. The dataset is included in the bnlearn package. To load the data, run data(“asia”). ...

#ARIMA #ARMA

Time Series Cheatbook

1 Library 1.1 Generate two time series and use smoothing filter 1.2 Casuality and Invertiblity. 1.3 ACF and Theortical ACF 2 Visualization, detrending and residual analysis of Rhine data. 2.1 ACF Plot 2.2 Detrending using linear regression 2.3 Detrending using kernel smoother 2.4 Detrending using seasonal means model 2.5 Model tuning using SetpAIC 3 Analysis of oil and gas time series 3.1 Checking Stationary 3.2 Log transformation to fix stationary 3. ...

#Machine Learning #Cheatbook

Machine Learning CheatBook

1 Simple Tasks 1.1 Library 1.2 Reading Excel 1.3 Spliting the Datasets 1.3.1 Divide into train/test 1.3.2 Train/test/validation 1.4 Custom code for Cross-Validation 1.5 Misclassification error calculation 1.6 Assume that mortality y is Poisson distributed, where Y!=12..n . Write an R code computing the minus-loglikelihood of Mortality values for a given lambda. Compute the minus log-likelihood values for lambda=10,110,210,…,2910 and produce a plot showing the dependence of the minus log-likelihood on the value of lambda. ...

#Visualization #ggplot2 #Cheatbook

Visualization CheatBook

1 Reading Data 2 Data Mugging 2.1 Quantile Computation 2.2 Scaling the Data 2.3 Distance Matrix between rows 2.4 Non-metric MDS 2.5 Principle Component Analysis 2.6 Types of Projection 2.7 Types of easing 2.8 Sorting dataset 2.9 Colour selection palette 2.9.1 Adding custom colours without palette 3 Single Plots 3.1 Density Plot 3.1.1 Density Plot with Outlier Highlight using GGplot2 3.1.2 Density Plot with Outlier Highlight using Plotly (converting from ggplot2) 3. ...

#bayesian statistics

Bayesian Statistics

1 Introduction 1.1 Frequentist vs. Bayesian View 1.2 Probability Distribution 1.3 Marginal Probability 1.4 Conditional Probability 1.5 Bayes Rule 2 Cheatbook 3 Lab1 Question 1: Bernoulli … again. 3.1 a) Draw random numbers from the posterior \[\theta|y \sim \beta(\alpha_0 + s; \beta_0 + f), y = (y_1,y_2,y_3....y_n)\] and verify graphically that the posterior mean and standard deviation converges to the true values as the number of random draws grows large. ...

#Deep Learning #Neural Network

Neural Networks

1 Introduction 1.1 Adding More Complexity Layer by Layer 1.2 Implementation Structure 1.3 Different Networks for different tasks 1.4 Why is Deep Learning taking off now? 2 Logistic Regression 2.1 Loss function of Logistic Regression 2.2 Loss function over all cases or Cost function 2.3 Gradient Descent 2.4 Applying gradient descent to logistic regression 2.5 Final Results 3 Building a Neural Network 3.1 Activation function 3.2 The need for non-linear activation function 3. ...

#K-Means #PAM #ROCK

Clustering

1 Introduction 1.1 Distances used in clustering 1.1.1 Distance between the objects 1.2 Types of variables 1.3 Calculating distance based on types of variables 1.3.1 Binary variables 1.3.2 Categorical Variables: 1.3.3 Nominal variables and Ordinal Variables: 1.3.4 Mixed variables 1.3.5 Cosine Similarity 2 Clustering Methods 2.1 Partitioning Method 2.2 Hierarchical Method 2.2.1 Agglomerative Approach 2.2.2 Divisive Approach 2.3 Density-based Method 1 Introduction Cluster is a group of objects that belongs to the same class. ...

#Linear Congruential Method #Inverse CDF Method #MCMC

Computational Statistics

1 Libraries 2 Distributions 2.1 Relationship between distributions 2.2 Bernoulli Distribution 2.3 Beta Distribution 2.4 Exponential distribution 2.5 Pareto Distribution 2.6 Uniform Distribution 2.7 Normal Distribution 2.8 Extra suggestions 3 Random Sampling from Uniform distribution 3.1 Sampling based probabilities proportional to the number of inhabitants of the city 4 Numeric Precision 5 Random Number Generation 5.1 Implementing the Varience 6 Scaling to get better results 7 Split the data into train and test 8 Loess Model with brute force of finding in minimizing of a function 9 Optimize function to find the minimum 10 Optim function to find the minimum 10. ...