fbpx

Time Series And Forecasting

Decomposition and Forecasting Methods

Ad-Hoc Course Registration:

  • Date: 1 – 4 February 2021
  • Time: 18.30 – 21.30
  • Venue: Menara Kadin Lantai 4, Jl. H. Rasuna Said, Jakarta Selatan
  • Investment: Rp. 5.200.000
  • Date: 1 – 4 February 2021
  • Time: 18.30 – 21.30
  • Investment: Rp. 2.600.000

REGISTER

Course details :

Decomposition of time series allows us to learn about the underlying seasonality, trend and random fluctuations in a systematic fashion. In this workshop, we learn the methods to account for seasonality and trend, work with autocorrelation models and create industry-scale forecasts using modern tools and frameworks.

We strongly recommend that you complete the pre-requisite workshops prior to taking this course. Some concepts presented throughout the lecture may be less-than-ideal for practitioners who have not completed the pre-requisite courses.

Schedule

  • Working with Time Series

    Day 1

  • Time Series in Action

    Day 1

  • Classical Decomposition

    Day 2

  • Techniques to Work with Time Series

    Day 3

  • Learn-by-Building

    Day 4

Course Producer

Samuel Chan

An  RStudio-certified instructor and machine learning practitioner in the field of marketing automation, fraud detection, finance and e-commerce.  Samuel is Indonesia’s top-ranked Stack Overflow user in R (top 5% worldwide) for three years running, and boasts certifications from RStudio, Microsoft, MongoDB, Neo4J Database, Stanford University, John Hopkins University, among others.

Prior to Algoritma, he has 8 years of working experience, including a stint as in-house consultant to several public-trading companies from his time staying in China, Japan and Singapore. He is today an active trainer and consultant for various companies in the financial industry. He has guest lectured in various campuses: Binus, NUS (National University of Singapore)’s The Logistics Institute, University of Indonesia, Universitas Gadjah Mada (UGM), Binus, Institute of Technology Bandung (ITB), Telkom University etc. Courses he authored are offered also in Singapore through Ngee Ann Polytechnic.

Samuel is also among the first recipients of Microsoft Professional Program Certificate in Data Science in Southeast Asia, having demonstrated proficiency in R, Python, Microsoft Azure, SQL / T-SQL, PowerBI and a list of other technologies, and among the first to be certified in RStudio’s program. Technical committee member and competition judge on Finhacks 2018, the largest Machine Learning competition of the year organized by PT. Bank Central Asia (BCA) and DailySocial.

4-Day Workshop Modules

Syllabus: Time Series & Forecasting

Module 1: Time Series I


Working with Time Series

  • Application of Time Series
  • Definition of a ts Object
  • Functions to Work with Time Series

Time Series in Action

  • Indonesia’s Gas Emissions, 1970-2012
  • Frequency, Start and End
  • Time Series Plots

Classical Decomposition

  • Trend, Seasonality and Residuals
  • Understanding Lags
  • Additive vs Multiplicative

Classical Decomposition in Action

  • Monthly Airline Passenger, 1949-1960
  • The decompose Function
  • Understanding Smoothing

Techniques to Work with Time Series

  • Adjusting for Seasonality
  • Detrending
  • Decomposing Non-Seasonal Time Series

Module 2: Forecasting


Forecasting I

  • Simple Moving Average
  • Simple Moving Average from First Principles
  • Log-Transformation

Forecasting II

  • Forecasting Using One-sided SMA
  • Forecasting Using Exponential Smoothing
  • Holt’s Exponential Smoothing

Forecasting III

  • The beta and gamma Coefficients
  • Mathematical Details
  • Holt-Winters Exponential Smoothing

Advanced Time Series

  • ACF and PACF
  • ARMA and ARIMA Models
  • Stationarity and Differencing

Advanced Time Series II

  • Augmented Dickey-Fuller (ADF) Test
  • Seasonal ARIMA
  • Tips to Work With xts
  • Facebook’s Prophet
  • Quantmod for Quantitative Traders

Academy Modules


Graded Quiz

Learning-by-Building Module (3 Points)

Forecasting the Crime rate in Chicago

  • Download the dataset from Chicago Crime Portal, and use a sample of these data to build a forecasting project where you inspect the seasonality and trend of crime in Chicago. Submit your project in the form of an RMD format, and address the following questions:
    • Is crime generally rising in Chicago in the past decade (last 10 years)?
    • Is there a seasonal component to the crime rate?
    • Which time series method seems to capture the variation in your time series better? Explain your choice of algorithm and its key assumptions

The student should be awarded the full (3) points if they address at least 2 of the above questions.

Ad-Hoc Course Registration:

  • Date: 1 – 4 February 2021
  • Time: 18.30 – 21.30
  • Venue: Menara Kadin Lantai 4, Jl. H. Rasuna Said, Jakarta Selatan
  • Investment: Rp. 5.200.000
  • Date: 1 – 4 February 2021
  • Time: 18.30 – 21.30
  • Investment: Rp. 2.600.000

REGISTER

Workshop Receivables:

  • Workshop Lecturer’s Notes

    Including 2x Course Books (PDF), HTML files, course transcripts (if any).

  • Highly-accelerated Learning

    Learn under the assistance of mentorship of our lead instructor and a band of qualified teaching assistants throughout the 4-day course.

  • Certification of Completion

    Show current and prospective employers that you’ve completed the course with a signed certificate of completion.

  • Quality Learning Environment

    We pay meticulous attention to the logistical details of our workshops: quality audio and visual setups, comfortable sitting arrangements, small group size. Dinners are included for evening workshops.

  • Supplement Materials

    Receive supplement datasets to practice on, reference notes, working files (R Notebook or Jupyter Notebook), and other materials that will help you master the topics.

This workshop is recommended for:

The Time Series and Forecasting workshop is an intermediate-level programming workshop best suited to R programmers that are taking their first steps into data science and machine learning.

Students are assumed to have a working knowledge of R and have completed the necessary pre-requisites. Consider taking the pre-requisite course or a beginner-level course instead if you have no prior programming experience or statistics knowledge.

Past Workshops in this Series:

Students work through tons of real-life examples using sample datasets donated by our team of mentors and corporate partners. We believe in a learn-by-building approach, and we employ instructors who are uncompromisingly passionate about your growth and education.

Part of the Machine Learning Specialization

This workshop is part of the Machine Learning Specialization offered by Algoritma Data Science Academy. Participants are rewarded with a certificate of completion upon passing criteria, and are encouraged to advance further in the respective data science specialization.