Applied Time Series Analysis and Forecasting with R Projects

Explore a univariate time series of labor force participation for ages 25–54 in Singapore (1980–2007), imported via scan and prepared for Holt linear trend modeling.

Explore how to handle time series with trends using Holt linear and damped models and Arima, compare forecasts, and enhance visualizations beyond standard plots in R.

The exponential smoothing family offers models like simple, Holt, Holt-Winters, and ETS to fit time series with level and trend dynamics, with Holt linear trend for non-seasonal data.

Explore building a Holt linear trend model in R's forecast package for Singapore labor participation, including alpha, beta, confidence levels, and damped forecasts with the damping parameter.

Learn how Arima models capture time series structure using autoregressive, differencing, and moving average components in R, with auto.arima for univariate datasets and practical forecasting.

Explore time series visualization with ggplot2 and the forecast package, building an auto plot with auto layers for Holt, Holt damped, and ARIMA forecasts, plus legends and titles.

Compare in-sample forecasts to actual Singapore employment data using three models, visualize fitted values, and discuss why training-test splits still matter for model selection.

Analyze monthly Germany inflation data (2008–2017) to illustrate time series analysis: month-on-month rates, seasonality, model choices, and how negative values push additive exponential smoothing and ARIMA decisions.

Import the German inflation data from Excel into R by removing headers and forming a continuous vector, then convert it to a monthly time series with start=2008 and frequency=12.

Explore seasonal time series analysis for German monthly inflation rates, focusing on frequency, seasonal decomposition, ets and arima models, and cross-validation with tsxv to compare forecast accuracy.

Use seasonal decomposition to split a time series into trend, seasonality, and remainder with additive or multiplicative models. Learn about decompose, STL, STR, SDL, and forecast methods in R.

Explore seasonal arima in r using manual pdq with acf/pacf or auto.arima, compare non-seasonal and seasonal parts, and forecast, alongside ets as an alternative.

Demonstrate exponential smoothing with ets in r, comparing additive and multiplicative options, tuning alpha, beta, gamma, and phi, and generating a 60-period german inflation forecast.

Compare time series models using cross-validation with the Tsxv function to assess forecast accuracy beyond information criteria, favoring Arima over exponential smoothing based on MSE.

Scrape Novartis (NVS) data from Yahoo Finance and build a regular five-day series. Impute missing days using locf and compare ARIMA and ETS to identify buy or sell timing.

Learn to extract stock data from Yahoo! Finance in R using Quant Mod's get symbols function with multi-ticker queries and Google Finance as an alternative XTS time series.

Preprocess irregularly spaced data by regularizing into equal intervals to reveal seasonality. Compare Arima and ETS forecasts and examine ACF and PCF to assess autocorrelation and limited patterns.

Convert irregular Yahoo stock data into a regularly spaced time series by merging with a full calendar of trading days, imputing missing days, and exploring weekly seasonality in R.

Convert data to a five-day time series, plot seasonality, compare highs and lows by weekday, and apply median-based baselines with STL decomposition to reveal small but actionable patterns.

Apply state of the art time series analysis using neural nets and interactive charts to clean, model, and forecast restaurant revenue for 36 months, with interactive digraph visuals in Shiny.

Learn how to clean a messy time series dataset with missing values, outliers, and misformatted quotes, preparing monthly restaurant revenue data with seasonality and potential trend for analysis.

Clean a messy monthly revenue dataset by importing, removing quotes, separating columns, converting to a numeric time series, and handling missing values and outliers with forecast's clean function.

Fit a neural network autoregression model to a clean time series using lagged values and seasonal lags, and forecast 36 months with prediction intervals.

explore creating interactive time series visuals in R using the digraphs package, with range selectors, hover details, forecasts (36 months) from a neural net, and optional Shiny or Markdown deployment.

Master applied time series analysis with R projects, comparing exponential smoothing and ARIMA models on trend and seasonal data. Learn data preprocessing, cross-validation, and dynamic regression approaches.