Reference no: EM133696966 , Length: word count:1000
Artificial Intelligence and Machine Learning
Assessment - Individual Evaluation Activity (Evaluating Neural Network Models)
Individual Implementation and Problem Solving (Puzzle)
Your Task
Evaluate the predictive modelling capability of neural networks and other prediction models in the Orange Data Mining Application. The assessment will involve implementing predictive models using software and solving various machine learning problems encountered in business.
Assessment Description
There has been a recent advent of Neural Networks including applications in deep learning. Analytics professionals can run basic Deep Learning applications via the browser and no-code platforms as the algorithms use hardware accessed via the cloud to provide the required performance.
In this assessment, you will implement an image classification model using Orange Data Mining's image analytics toolbox. You will be provided with a dataset containing images.
Assessment Instructions
In class: Implementation. Use the Orange data mining software to perform both deep learning- based image classification and classical predictive analytics tasks.
Image Classification -Construct a predictive model using the Image Analytics widgets in Orange.
Analyse tabular data using Neural Networks (NN)- Construct a predictive model using NN and other widgets in Orange.
In class: Problem Solving (Puzzle). Based on the Orange output, solve the accompanying puzzles in the assessment sheet designed to test your understanding of various prediction models and general knowledge on machine learning.
Here are the things you'd want to revise in preparation for assessment 2:
-Decision trees: prediction, class probabilities
-Perceptron model: prediction, training
-Neural network: image classification, embedding, network complexity
-Logistic regression: prediction, class probabilities
-Naive Bayes: prediction
-Word vectors: computing word vectors
-Orange workflow: data loading, train-test split, model training, testing
-Confusion matrix: matrix construction, computing precision, recall