This case study aims to develop an interpretable regression model which could provide robust and reliable estimates of life expectancy from an optimal set of observations and predictors, while delivering accurate predictions when applied to new unseen data.

This presentation illustrates various penalized survival models for high-dimensional data in R.

This presentation illustrates various exploration and visualization methods for extracted dimensions obtained from principal component algorithms in R.

This presentation illustrates various sample size and power calculations for clinical research proportion comparison tests in R.

This presentation illustrates various sample size and power calculations for clinical research mean comparison tests in R.

This presentation illustrates various oversampling and undersampling algorithms for augmenting imbalanced data prior to model training in R.

This presentation illustrates various performance metrics which are adaptable to censoring conditions for evaluating survival model predictions in R.

This presentation illustrates various robust model variants applied to handle quasi-complete or complete separation during logistic regression modelling in R.

This presentation illustrates various density and distance-based anomaly detection algorithms for estimating outlier scores in R.

This presentation illustrates various isolation forest-based anomaly detection algorithms for estimating outlier scores in R.

This presentation illustrates various density-based clustering algorithms for identifying multivariate outliers in R.

This presentation illustrates various dichotomization thresholding strategies for optimally classifying categorical responses in R.

This presentation illustrates the implementation of the gradient descent algorithm in estimating the coefficients of a linear regression model in R.

This presentation illustrates various clustering algorithms for segmenting information in R.

This presentation illustrates various dimensionality reduction algorithms for extracting information in R.

This presentation illustrates various remedial procedures for handling skewed data with extreme outliers for classification in R.

This presentation illustrates the implementation of simulated annealing and genetic algorithms in selecting informative predictors for a modelling problem in R.

This presentation illustrates the implementation of univariate filters in selecting informative predictors for a modelling problem in R.

This presentation illustrates the implementation of recursive feature elimination in selecting informative predictors for a modelling problem in R.

This presentation illustrates various model-independent feature importance metrics for predictors with dichotomous categorical responses in R.

This presentation illustrates various model-independent feature importance metrics for predictors with numeric responses in R.

This presentation illustrates various cost-sensitive learning procedures for handling class imbalance in R.

This presentation illustrates various remedial procedures for handling imbalanced data for classification in R.

This presentation illustrates various evaluation procedures for hyperparameter tuning strategies and resampling distributions in R.

This presentation illustrates various predictive modelling procedures for multiclass categorical responses in R.

This presentation illustrates various predictive modelling procedures for dichotomous categorical responses in R.

This presentation illustrates various predictive modelling procedures for numeric responses in R.

This presentation illustrates various resampling procedures as applied during model hyperparameter tuning and internal validation in R.

This presentation illustrates the best practices when developing and evaluating prognostic models for clinical research using R.

This presentation illustrates various missing data pattern analyses, imputation method evaluations and post-imputation diagnostics in R.

This presentation illustrates the survival analysis and descriptive modelling steps for a three-group right-censored data with time-independent variables using the Cox Proportional Hazards Model in R.

This presentation illustrates the survival analysis and descriptive modelling steps for a two-group right-censored data with time-independent variables using the Cox Proportional Hazards Model in R.

This presentation illustrates various treatment comparison tests applied to data with a single factor variable (2-level) and a single response variable (numeric) in R.

This presentation illustrates various data quality assessment, preprocessing and exploration methods in R.

This presentation illustrates various data quality assessment, preprocessing and exploration methods in R.

This project summarizes the exploratory data analysis conducted for the pre-and post-structural topic modelling results involving the words, topics and documents obtained from the media contents of Philippine broadsheets.

This project implements the logit model under the binary response GLMs for understanding the factors driving the knowledge and awareness of survey respondents on immunization. Y corresponds to the measured knowledge on immunization. π represents the probability of having sufficient knowledge on immunization and n refers to the observation data for the 635 respondents used for the model.

This project evaluates the reliability of statements used to measure an abstract concept through likert scales. Median test was the primary tool used to identify the statements which effectively delineated responses from the survey respondents.

This exercise uses combined and separate ratio estimation in a stratified random sample to estimate the population total of a certain variable.

This exercise uses ratio estimation in a simple random sample to estimate the population mean of a certain variable.

This exercise generates the ANOVA table for simple and stratified random sampling methods.

This exercise estimates the population mean using simple and stratified random sampling.

This exercise computes the T2-statistic as used in testing hypotheses on linear combinations of the elements of the mean vector μ. Under the general linear hypothesis (GLH), the interest is in testing the hypotheses as follows: (1) Ho : Aμ = b and (2) Ha : Aμ ≠ b

This exercise implements simultaneous inference for means using the Roy-Bose simultaneous confidence intervals and the Bonferroni method of multiple comparisons.

This exercise implements simple analysis for Poisson point patterns as applied on spatial data.