AI for Junior
Artificial Intelligence (AI) and Machine Learning have made significant breakthroughs in the past decade. AI has deeply impacted our lives through its application in online search, mobile devices, social networks, and image analysis. The goal of this course is to provide a introduction to AI for high school students. This course will walk students through the definition of AI, the history of AI, AI algorithms, and the application of AI in real life situations.
1. High school students who want to learn AI before college.
- Lecture can be attended in two ways (Live and Playback video)
- Live Lecture schedule:
- Lecture will be on live every Tuesday and Thursday at 8pm(EST).
- First lecture begins on Nov. 29th 2018 at 8pm(EST).
- Each lecture is about 2 hr.
- Two lectures a week.
- If any lecture schedule changes, a notification email will be sent to all enrolled users.
- PlayBack Lecture Video:
- Playback recording video will be available after each live lecture.
- Playback recording video can be viewed unlimited times within one year after enrolment.
- Lecture Slides:
- Lecture slides will be available after each live lecture.
Owen Jian got his master of Applied Science in Quality System Engineering at Concordia University. After that, he has been focusing on Machine Learning and Deep Learning for years in industry, and obtains a strong knowledge and understanding about AI. His research mainly focuses on AI training model such as CNN, RNN and LSTM. Owen also has quite a lot software industry experience in Air Canada, Rogers. Currently he works as a software developer in Ceridian.
Isha Shingari was an Assistant Professor in both Institute of Innovation in Technology & Management of India and Jaipur National University of India. She is currently pursuing her PhD degree in Computer Science. Her search focuses on core expertise of Training model, data analysis and pre-processing of data. Establishing scalable, automated process for large scale data analyses, neural networks and model validation are also within her search.
Yixin Zhang, a PhD at University of Alberta, is also a Data Scientist and he is currently working at University of Alberta as well. Yixin works on large data analysis on oil sends, visualization and prediction on oil sands data for further processing. He also works on building probabilistic models of complex which utilizing Bayesian and monte Carlo methods.
- Defining Artifical Intelligence
- AI Background and History
- AI Hardware
- Comtemporary AI Influencers and their works
- Challenges and Opportunities
- AI Research
- AI Application and Scenarios
- AI Trends
- Machine Learning and AI
- Fundamental Concepts behind AI
- AI Ecosystem and Terminology
Introduction to AI Systems
- Image Recognition
- Video Identification and Classification
- Audio, Speech, and Translation ("Chatbot")
- Sentiment Analysis
- Music Maker
- Self-driving Cars
- Image and Video Inference
- Reinforcement Learning
- Gaming (Chess, Online Games)
- Recommendation System
- Knowledge Graphs
- Blockchain - AI Tracking
- Expert Decisions
Introduction to AI Skill Sets
- Data Structures, Types, and Managment
- Data Sample and Bias
- Data Cleaning
- Features and Extraction
- Features Engineering and Transformation
- Dimension Reduction
- Classification and Regressino
- Aritifical Neural Networks
- Natural Language Processing
- Deep Learning
- Reinforcement Learning
- Defining Artificial Intelligence
- AI Background and History
In 1956, AI was founded as an academic discipline, followed by optimism of early AI between 1960 and 1970.
Minsky wrote the book 'Perceptron'. Followed by two stages of 'AI Winter' in 1974 - 1980 (memory and speed constraints) and 1987 - 1993 (economic bubble) that saw significant reductions of AI funding.
Rebranding as 'Deep Learning' (2006), Hinton introduced the idea of unsupervised pre-training and deep belief nets using restricted Boltzmann machine.
The Breakthrough (2012), with Geoff Hinton's application of deep learning into Large Scale Visual Recognition Challenge (LSVRC), the error rate is reduced to 16%.
Around 1995 to 1997, AI and Deep Learning began to gain popularity with the appearance of large, high-quality labeled datasets, along with the availability of high-performance GPUs.
Amazon's recommender system can identify the users who are interested in purchasing based on customer's online behavior. That is why Amazon can post many new products before we even know about them. The interested thing is that when we browse the website and purchase some products, the recommended products are seemingly irrelevant to the products we are buying.
Imagine a robot with AI can help teachers with reviewing knowledge for students and grading the exams? With AI, the knowledge delivered from teachers can be stored, learned and even processed by the robot student. With its knowledge refined more and more with each passing year while sitting together with real students listening to lectures, the robot will finally develop and become more intelligent, it can be successfully used to communicate with students as teaching assistants. At the same time, since AI is able to store astronomical number of study cases and students' knowledge level and scope, it can quickly organize many versions quiz and exams including the solutions.
If you are not sitting in a driverless car for traveling every day, you have no idea what you're wasting - time! Time is money! Average commuter times in city areas is around 30 minutes, Image almost 200 hours saving for each person, how much dollar value can be created? With humans no longer driving, AI in driverless car is likely to save up to an hour every day – time that will undoubtedly have many spin-off benefits from wellbeing to boosting the economy. What is more, the self-driving cars can reduce can reduce traffic congestion by eliminating stop-and-go waves caused by human driving behavior, this is because AI enables driverless vehicles to communicate with each other and their surroundings, hence they are able to identify the optimal route. This helps spread demand for scarce road and speed. The more important benefit of such AI-controlled car has the potential in the future to reduce deaths and injuries from car crashes, particularly those that result from driver distraction.
It is well known that AI is now applied to automatic speech recognition (ASR): speech-to-speech and voice-to-text translation. Many translation AI products is to get this to 95 percent accuracy. A particular example is the Google's Duplex, a new technology for conducting natural conversations to carry out ͞real world͟ tasks over the phone. For instance, you wish to book a restaurant. The humans on the other end of the line were actually Duplex that handled all challenges with ease, like when the restaurant said it has vacancy until 1 hour after. The core of Duplex is a recurrent neural network (RNN) designed to cope with speech translation, built using TensorFlow Extended (TFX). To obtain its high precision, the AI trains Duplex͛s RNN on a corpus of anonymized phone conversation data. The network uses the output of Google͛s
automatic speech recognition (ASR) technology, as well as features from the audio, the history of the conversation, the parameters of the conversation (e.g. the desired service for an appointment, or the current time of day) and more.
You must be enrolled to this course before you can provide a review.