• Data scientist at a tech company in Indonesia • Google Developer Expert in Machine Learning • Co-host podcast Kartini Teknologi (kartiniteknologi.id) @galuhsahid Hi! I’m Galuh.
Outline @galuhsahid • What is machine learning? • What are the differences between AI, ML, deep learning…? • Examples of ML/AI implementation • How to implement ML/AI? • Workflow • Tools & resources • Challenges in ML/AI implementation
if pixel[5][7] is black and pixel [5][6] is black and pixel [5][8] is black and …: if pixel[6][7] is black and pixel[6][7] is black and …: return “panda” … … … else: return “cat” Photo by Damian Patowski from Unsplash @galuhsahid
if pixel[5][7] is black and pixel [5][6] is black and pixel [5][7] is black and …: if pixel[6][7] is black and pixel[6][7] is black and …: return “panda” … … … else: return “not cat” Photo by Dušan Smetana from Unsplash @galuhsahid
@galuhsahid Define your machine learning problem Acquire & prepare your data Train your model Serve your model Adapted from Introduction to ML Problem Framing Explore your data Workflow Implementation
Step #2 Acquire & prepare your data @galuhsahid You need to know: • What are the types of data that you can use • Where to get them • How to get to know your data • How to prepare your data
Exploratory Data Analysis Getting to know your data - Analyze your data to summarize their main characteristics - Examples include: check for basic statistics (e.g. mean, median), missing data, outliers @galuhsahid
Misconceptions about AI/ML Challenges @galuhsahid Common myths: • The most important part in implementing ML is building models • Hyperparameter tuning is the only way to improve our model • We can get a 100% accuracy • ML is the answer for everything
Data availability Challenges @galuhsahid • ML needs data! • However, since the awareness of data in the industry is relatively new, sometimes companies want to jump to implementing AI/ML without thinking about whether they actually have the data or not • What about open data? Indonesian open data still needs lots of improvement https://hackernoon.com/the-ai-hierarchy-of-needs-18f111fcc007
Data infrastructure readiness Challenges @galuhsahid • Before implementing AI/ML, we need to have the data infrastructure ready first. Building data infrastructure needs planning and takes time. https://hackernoon.com/the-ai-hierarchy-of-needs-18f111fcc007
Supporting resources Challenges @galuhsahid • Sometimes the success of an ML project also depends on the resource available • Example: NLP resources for Indonesian are limited compared to English https://github.com/kmkurn/id-nlp-resource
More machine learning • On building ML projects: First Steps Towards Your First Machine Learning Project • On ML with JavaScript: Machine Learning on the Web • On ML with TensorFlow: A Whirlwind Tour of Machine Learning with TensorFlow @galuhsahid