This course introduces the basic concepts of Artificial Neural Network, its applications and techniques for developing models to predict outputs for any specified input dataset. The basic principles shall be demonstrated in a very lucid manner on an excel worksheet. Moreover, it is not essential for participants to have any prior knowledge regarding the subject.
Course Structure: 10 modules of 2.5 hours each, delivered over 5 days Each day will consist of 2 modules of 2.5 hours, with multiple breaks.
Course Level: Foundation Duration: 5 days Instructor: Mohit Narain
Designed for you, if you are...
A professional handling data in teams such as operations, technical services, safety, geology, research, reservoir engineering and data management.
How we build your confidence
Highly interactive course
Practical oil & gas case examples will be presented and adequate exposure will be provided to the participants through exercises
Every participant will be able to hone his skills on several input datasets.
The benefits from attending
After completing the course, you will be able to:
Manage raw data more effectively.
Develop basic feedforward neural network models on excel worksheets.
Participate in discussions related to applications of neural modelling.
Expose other participants to new technology for further development.
Attend advanced courses on the subject.
By the end of the course you will feel confident in your understanding of:
Fundamental Structure of the Artificial neural network model.
The concept of backpropagation technique.
The significance of Activation functions.
Use of the excel worksheet for minimizing cost functions.
Applications of the neural model
Similarities to biological neurons
Concept and structure of Artificial Neural Network
Managing the data for a neural network
Linear and non-linear separability
Commonly used activation functions
Effect of parameters on the activation functions
Utilising activation functions for mapping outputs - forward pass for a small data set
Determining the cost function on a small data set based on activation functions
Procedure for backpropagation to develop a new set of weights
The local & global Minima-Avoiding the local minima using the momentum term
Minimizing the cost function using the Excel worksheet and finalizing the weights for the model
Concept of training, validation and test sets.
Comparing the cost functions for the validation and test datasets based on the weights obtained for the training dataset
Effect of increasing the nodes in the hidden layer on minimizing the cost function
Adding a hidden layer to the model and its impact on minimizing the cost function.
Estimating the cost function for the validation and test datasets based on the model obtained from training dataset
The concept of Multiple Input Multiple Output (MIMO) and Multiple Input Single Output (MISO) models
Analysis of weights for MIMO & MISO models using different activation functions
Comparing minimized cost functions for validation and test datasets for the MIMO model & both MISO models
Developing the model for compressor liquid and gas data example with two different activation functions using two inputs - a single hidden layer and two outputs
Compare the effect of adding a hidden layer on the training, validation, test data in above exercise
Review the graphical correlation between the true and model predicted values on the test data