Bootcamp on Data Science using R language

Extract insights from data by collecting, cleaning, analyzing, and presenting to inform decisions. Emphasize understanding the problem, modeling with machine learning and deep learning, evaluation, deployment, and automating data workflows.

Define data as raw facts that support a phenomenon, like a fever reading. Explore structured and unstructured data, and qualitative and quantitative types—nominal, ordinal, discrete, continuous.

Explore the growing data science landscape, roles like data engineer, data analyst, data architect, and data scientist, and the data discovery, cleaning, modeling, and tuning steps essential for data-driven decisions.

Explore data science tools from Hadoop MapReduce for big data analysis to Python libraries for machine learning and visualization, with RapidMiner, Tableau, Power BI, and KNIME for preparation and reporting.

Understand the data science process flow from defining the problem and locating data to cleaning, exploring with visualization, building a model, and interpreting results with dashboards.

Discover how data science drives healthcare insights, aviation revenue management, and manufacturing optimization through predictive analytics and fault diagnostics, while enabling weather forecasting and business intelligence.

Discover the R language and RStudio as your tools for statistical computing and graphics, leveraging CRAN packages for data analysis, time series, classification, clustering, machine learning, and deep learning.

Learn how to install the R language and the RStudio IDE across Windows, Mac, and Unix, including downloading, selecting defaults, and verifying installation.

Explore the RStudio interface and tools, including console, environment, history, and editor, to write and run R scripts. Learn to manage variables, install packages via cran, and customize appearance.

Set your working directory to a designated data science folder inside the R directory to organize files and simplify importing, exporting, and manipulating data.

Explore core data types and variables in R, including integers, numeric, complex, character, and logical, set a working directory, create and inspect variables, and perform type checks and conversions.

Explore arithmetic operations in r by declaring variables, using sum for totals, printing results, and tracking salary, expenses, and savings within a clean environment.

Learn to create and inspect data frames in R, set a working directory, clear the environment, build a frame with id, name, salary, and joining date, and use str.

Master the data science methodology from defining research questions to collecting, cleaning, and exploring data, then modeling, presenting insights with dashboards, and building automated data pipelines.

Distinguish primary and secondary data collection techniques, applying surveys, interviews, observations, experiments, and web scraping to study unique problems and existing phenomena.

Learn how web scraping automates data collection from websites, handles semi-structured data, builds pipelines, and respects robots.txt and robots exclusion protocol for privacy-aware analyses via APIs.

Improve data extraction with a hands-on web scraping workflow in R language, using selector gadget to identify CSS selectors, read HTML, and build a CSV of IMDb top movies.

Learn how data pre-processing sanitizes raw data, reducing garbage in and garbage out, by cleaning, standardizing, encoding categorical data, scaling, and validating data to improve model quality.

Set up a data preprocessing workflow in R, organize a folder, import the IMDb ratings data, and clean year and duration while preparing for training and testing set and scaling.

Learn how to handle missing data in R by imputing with the mean, and perform data preprocessing steps like creating new columns, dropping redundant columns, and reordering features.

Discover how to handle categorical data by converting nominal and ordinal categories into numeric codes using as.factor and factor, turning true/false into 1/0 with labeled levels.

Split a data set into training and test sets in R using the CA Tools sample.split function with an 80/20 ratio, and create corresponding training and testing subsets.

Scale data in R using the scale function to center and adjust a column, preserving original data while preparing heterogeneous data for distance-based models like SVM and KNN.

Explore how statistics reveal data reality through uncertainty and variation, guiding informed decisions in data science. Understand population vs. sample, sampling, descriptive and inferential statistics, and confidence intervals.

Explore descriptive statistics for data science: summarize data with central tendency, dispersion, and shape using mean, median, mode, range, standard deviation, variance, skewness, and kurtosis, with histograms and box plots.

Learn to compute mean, median, and mode in R using a near-earth objects data set, including installing and using the mode est library to calculate mean diameter, velocity, and distance.

Explore data dispersion by calculating standard deviation and variance in R using a near Earth objects dataset, including setting the working directory, importing data, and interpreting results.

Explore kurtosis and skewness using the moments library in R, apply them to the Near Earth objects minimum distance data, and interpret distribution shape and outliers.

Explore inferential statistics by contrasting it with descriptive statistics, and learn how samples estimate population parameters, test hypotheses, and construct confidence intervals.

Explore how data distributions inform inference, grasp the central limit theorem, and learn to compute confidence intervals using the z distribution with a height example.

Learn to compute a 95% confidence interval for iris sepal length in R using a z score, including the sample mean, standard deviation, and the z critical value.

Discover the student t distribution and its heavier tails. Learn to use the t distribution for confidence intervals and hypothesis testing with small samples and unknown population standard deviation.

Explore confidence intervals and t tests in R with the iris data, focusing on sepal length, sample size, and the difference from z tests.

Explore hypothesis testing, including null and alternative hypotheses, significance level, p values, and rejection regions. Differentiate one-tailed and two-tailed tests and grasp interpretation with type I and type II errors.

Explore hypothesis testing in R with one-sample and two-sample t tests, null and alternative hypotheses, t statistics, p-values, and the Lebanese test for variance to compare groups.

Learn predictive analytics by applying historical data, statistical modeling, data mining, and machine learning, including linear and logistic regression, to forecast future outcomes and guide business decisions.

Explore how simple linear regression predicts a dependent variable from one independent variable using the regression line y = mx + c and ordinary least squares.

Build a simple linear regression model in R using GRE score to predict chance of admit, including data import, training and testing split, lm fitting, prediction, and ggplot2 visualization.

Explore how multiple linear regression uses many predictors to model a single output and estimate coefficients, with feature engineering illustrated by mobile price prediction.

Build a multiple linear regression model in R using the admissions dataset, from data prep and train-test split to fitting lm and evaluating predictions with mse, rmse, or mae.

Classify inputs into discrete classes and contrast with regression. Apply models like logistic regression, SVM, decision trees, Naive Bayes, boosting, and random forests to tasks such as spam detection.

Explore logistic regression, a binary classification method using the sigmoid function to predict outcome probabilities and convert them to 0 or 1, with binomial and multinomial variants, implemented in RStudio.

build a logistic regression classifier in r using UCLA data with GRE, GPA, and rank to predict admission; split data into training and testing sets, and evaluate with confusion matrix.

Learn how random forest classification forms an ensemble of decision trees that vote to predict outcomes, using bagging and bootstrapping to improve accuracy.

Build a random forest classification model in R to predict diabetes using the Pima Indians dataset, including data prep, training and testing splits, and evaluating with a confusion matrix.

Explore dimensionality reduction by reducing features and removing noisy or redundant dimensions to combat overfitting, improve accuracy, and speed up training, with techniques like feature selection and extraction, including PCA.

Explore principal component analysis, a dimensionality reduction method that standardizes data, builds a covariance matrix, derives eigenvectors and eigenvalues, and transforms data into uncorrelated, high-variance components.

Develop a PCA-driven classification workflow in R using iris data: reduce to two components, build an SVM with linear kernel, split data, scale features, and assess with a confusion matrix.