Introduction to Artificial Intelligence

• Four main definitions of artificial intelligence
• Artificial intelligence research and applications
• Building artificial intelligence systems
• Intelligent agent types
• intelligent agent task environments
• Observable, partially observable, and unobservable environments
• Deterministic and stochastic environments
• Levels of certainty in an environment
• Types of environmental behaviour

• Techniques used to solve adversarial problems
• Representing an adversarial problem
• Using the minimax algorithm
• Using Alpha-beta pruning to improve algorithm performance
• Evaluation functions
• Using cutoffs to perform adversarial searches
• Looking up tables to improve an agent’s performance
• How to play the game of chess
• Expect minimax values in stochastic games
• Evaluation functions used to search in a stochastic game
• Using monte carlo simulations

• Introducing CSPs
• Constraint satisfaction problems
• Search problems
• Constraint satisfaction algorithms
• Search algorithms
• Using a backtracking search
• Solving a constraint satisfaction problem
• Ordering variables
• Performing a backtracking search
• Arc consistency
• Other types of constraint consistency
• Solving a constraint satisfaction problem
• Constraint propagation
• Using the backjumping and forward checking inference method
• Local search algorithms
• Representing a Sudoku puzzle

• Defining NLP
• NLP applications and methods
• NLP operations (Regex, tokenization, and stemming)
• Porter stemming algorithm used to stem English text
• Performing entity recognition
• Basic models
• Building NLP models
• Text classification
• Naïve Bayes classification algorithm
• Retrieval techniques
• Parsing using NLP
• Machine translation
• Computer methods used to recognise speech

• How AI learns
• Different types of machine learning
• Choosing attributes to learn a decision tree
• Entropy and information gains
• Decision tree attributes
• Neural network types and structures
• Individual neurons
• Solving problems with neural networks
• Machine learning with a perceptron
• Multilayered neural network
• Convolutional neural networks
• Recurrent neural networks

• AI strategy expectations and buy-ins
• Challenges surrounding the adoption of AI
• AI implementation training
• Data and algorithms roles
• Planning and develop AI capability
• Challenges facing management when developing AI solutions
• Pitfalls of AI
• Organisational AI strategy elements
• Issues surrounding data quality, training, overfitting, and bias
• Assessing AI needs and tools
• discuss the various aspects of AI an organisation needs to address to plan for AI

• Reinforcement learning
• Additive rewards and discounted rewards
• The Bellman Equation
• Temporal difference learning
• Direct utility estimation
• Active and passive learning
• Exploration and exploitation in active reinforced learning
• Exploration policies used in learning algorithms
• Implementing and parts of Q-learning
• On-policy and off-policy learning
• Building function approximations

• Search problems used by AI agents
• Problems used for searching algorithms
• Brute force searching
• Breadth-first search algorithm
• Depth-first search algorithm
• Depth-limited search
• Iterative deepening search algorithms
• Best-first informed searching
• Heuristics properties
• Creating a heuristic function
• A* search algorithm
• Hill-climbing search algorithm
• Simulated annealing search algorithm

• Alexa developer console skills
• Using invocations in Alexa developer consoles
• Creating custom and built-in intents
• Using utterances and slots in Alexa developer consoles
• Building a Lambda function
• Alexa simulator skills testing
• Echosim skill testing
• Configuring a skill to use DynamoDB for persisting session data
• Creating an Alexa skill that takes advantage of multi-stage dialog

• Creating an agent for a chatbot in Dialogflow
• Default and custom intents in Dialogflow
• Dialogflow developer and system entities
• generate developer entities to extract information from user conversations in Dialogflow
• Training phrases in Dialogflow
• Actions and parameters in Dialogflow
• Static responses a bot can respond to
• Enabling the Small Talk feature for a chatbot
• Testing functionality of a bot
• Inline cloud functions in a Dialogflow
• Create a chatbot in Dialogflow

• Linear and non-linear human/chatbot conversations
• Input and output contexts
• Follow-up intents
• Entry point for non-linear conversations
• Chatbot dialog contexts
• Integrating Dialogflow chatbots with other platforms
• Deploy a fulfillment in Dialogflow
• Using actions on Google in Dialogflow
• Testing chatbot using Google Assistant
• Integrating Dialogflow chatbots with Google Assistant
• Chatfuel bot development platform building blocks
• Pre-built flows in Chatfuel
• Text and typing elements in Chatfuel
• Linear and non-linear dialogs

• Creating and using chatbots
• Classifications of chatbots
• Conversational flow of typical chatbot/human interface
• Building blocks for a typical chatbot built on Dialogflow
• AWS developer accounts
• Components of the Alexa Development Console
• Configuration of an AWS Lambda service
• Creating a developer account on Google’s Dialogflow
• Components of a Dialogflow developer console
• Chatbot use cases and technology stack

• Primary goals of machine learning and artificial Intelligence,
• Primary goals of deep learning and reinforcement learning
• Essential features of artificial intelligence
• Java development environment
• demonstrate the use of machine learning algorithms in Java
• Various implementations of AI
• Features and capabilities afforded by Deeplearning
• Configuring neural networks using DL4J
• Implementing artificial intelligence
• Predictive modelling with relatable algorithms
• Creating a Maven project
• Configuring a Weka library
• Creating a dataset for Weka

• Expert system languages
• Creating rule based expert systems with Jess
• Working with expert system shells using Java
• Recognising data notations
• Types of datasets
• Types of outliers
• Feature relevance search and evaluation techniques
• Principal component analysis data transformation
• Clustering implementation algorithms
• Implementing hierarchical clustering
• Graph modelling
• Using datasets with clustering

• Machine learning and artificial intelligence
• Classifications of machine learning algorithms
• Implementing KNN algorithms
• Implementing decision tree and random forest
• Linear regression analysis
• Implement Gradient boosting algorithms using Java
• Implementation of logistic regression using Java
• Probabilistic classifiers for statistical classification
• Implementing Naïve Bayes classifier using Java
• Using the K-Mean algorithm in ML applications

• Concepts and layers of neural networks
• Practical implementation of a simple neural network using Java
• Types of neural networks
• Implementing Hopfield Neural Networks
• Implementing back propagation neural networks using Java
• Types of activation functions in neural networks
• Types of loss functions
• Activation functions and loss functions using DL4J
• Working with hyperparameters in neural networks
• Practical implementation of Neuroph framework
• Arbiter hyperparameter optimisation library
• Deep learning concepts and components
• Differences between deep learning and graph model

• Multilayer networks and computation graphs
• Features and components of NLP
• Language and sentence detectors
• Utilisation of Tokenizer and Name Finder in NLP
• Detecting parts of speech to assign tags to words and sentences
• Classifying text and documents using the NLP model
• Implement recogniser, synthesiser and translator using Java

• Encoding data
• Autoencoders
• Principal component analysis
• Using principal component analysis with scikit-learn
• Applying autoencoders
• Fashion MNIST dataset for dimensionality reduction
• Autoencoders and their use cases

• Visual cortex with a neural network
• Applying a convolution to an input matrix
• Using scikit-image to read in an image
• Using a convolutional kernel with a convolutional layer
• Convolutional layers
• Pooling uses
• Using hyperparameters
• Structure of a convolutional neural network
• Overfitted models and the bias-variance trade-off
• Regularisation, cross-validation, and dropout
• Using CIFAR-10 dataset for image classification
• Splitting the dataset into training and test images
• Creating placeholders
• Convolutional and pooling layers
• Running training and prediction on the CIFAR-10 dataset
• Defining the role of convolutional and pooling layers in a convolutional neural network

• Machine learning algorithms
• Training and prediction phases in machine learning
• Traditional machine learning and deep learning
• Using TensorFlow for machine learning
• Installing TensorFlow based on the user’s environment
• Installing TensorFlow and work with Jupyter Notebooks
• Building and running a computation graph
• Using a TensorBoard
• Build and execute a computation graph
• TensorBoard variables and placeholders
• Working with feed dictionaries
• Grouping computations with named scopes
• Eager execution for prototyping and development
• Concepts of machine learning and TensorFlow
• Execute computation graphs

• Characteristics of supervised and unsupervised learning algorithms
• Applying unsupervised learning
• Objectives of clustering algorithms
• K-means clustering to group data process
• Performing and implementing k-means clustering
• Installing TensorFlow and work with Jupyter notebooks
• Generating random data for clustering algorithms
• Iris dataset of flowers
• Performing clustering and classification on the Iris dataset

• Load and explore training data
• Pre-process data to feed into the neural network
• Creating unique identifiers
• Creating a neural network for sentiment analysis
• Training the neural network for prediction
• Pre-process data to feed into the neural network
• Creating a lookup table to map words
• Training data sentences in the form of word identifiers
• Performing sentiment analysis using GloVe embeddings
• Using RNNs for sentiment analysis

• Linear regression problems
• General machine learning problems
• Model parameter training
• Loading a dataset (features and labels)
• Linear regression model data
• Building a base model
• Creating placeholders, training variables, and instantiate optimisers
• Train model parameters and visualise the result with Matplotlib
• Loss and summaries on TensorBoard
• High-level Estimator API
• Training a regression model
• Evaluate and predict housing prices using estimators
• Classification problems
• Logistic regression for classification
• identify data as being a continuous range or comprised of categorical values
• Training and test data to predict heart disease
• Training the high-level estimator for classification
• Binary classification machine learning models

• Text to numeric conversion using one-hot encoding
• Generate word embeddings
• Pre-trained models for word vector embeddings
• Working with recurrent neurons
• Construction of recurrent neural networks
• Training a recurrent neural network
• Long memory cells with the normal recurrent neuron

When you study with Australian Online Courses, you will be assessed using a competency-based training method.

Competency-based training focuses on the achievement of skills and knowledge against set criteria to ensure your competency is industry relevant. You will not be benchmarked against other students.

If you do not achieve a competency result on your first attempt, you have two more attempts to pass your assessment. So, you have three attempts in total to obtain a competency result.

In this way, you can complete your course in your own time and at your own pace with the assistance of unlimited tutor support.

In this course, you will be assessed via multiple-choice questions to determine your mastery of details and specific knowledge gained during your studies to achieve a ‘competent’ or ‘not yet competent’ result.

Benefits of Multiple-Choice Assessments

• Appropriate for assessing students’ mastery of details and specific knowledge.
• Can be used to assess both simple knowledge and complex concepts.
• Questions can be answered quickly to accurately assess a students’ mastery of many topics relatively quickly.
• Assessment can be quickly and reliably scored to achieve a ‘competent’ or ‘not yet competent’ result.
• As the answers are visible, multiple-choice questions offer the opportunity for the continuation of the learning process, offering educational value.

• Professional development that is widely recognised and respected;
• Improve your employment opportunities;
• Study online, anywhere via our elearning system;
• High-quality professional development programs written by industry experts;
• No entry requirements;
• All course materials provided online – no textbooks to buy;
• Unlimited tutor support;
• Easy to understand course content;
• We offer twelve (12) months’ access, with extensions available upon application (fees apply);
• Certificate of Attainment/Completion issued for your CV (hard copy provided); and
• Course may be tax deductible; see your tax advisor.

The approximate study hours for this course is 25 hours. We offer twelve (12) months’ access, with extensions available upon application (fees apply).

While printed materials are not available for purchase, transcripts of video lectures and other learning resources are available for printing via our elearning system.

You can start within 60 minutes during business hours when you enrol and pay in full with a credit card!

Credit card: Within 60 mins during business hours.

BPAY: Within 1-2 working days.

Internet Banking: Within 1-2 working days.

Cheque/Money Order: Upon receipt of mailed cheque.

Yes! You will receive a Certificate of Attainment upon successful completion of your assessment.

No. All courses are delivered online via our e-learning system and there are no work placement requirements in this course.

Unlimited tutor support is available throughout your studies via email during business hours Monday to Friday. Our Administrative team are available Monday to Friday via email, live chat and telephone.

Yes! We accept enrolments from individuals both within Australia and internationally; location is no barrier to entry into our programs.