Forecasting CO2 emissions using Deep Learning

Learn how to apply Deep Learning using Python for generating forecasts. Step by Step.
Forecasting CO2 emissions using Deep Learning
File Size :
1.74 GB
Total length :
3h 45m

Category

Instructor

Dr. Spyros Giannelos

Language

Last update

2/2023

Ratings

5/5

Forecasting CO2 emissions using Deep Learning

What you’ll learn

This course is part of the ENERGY-COMPLETE dot com official program.
Learn a 10-step Deep Learning – based methodology for producing forecasts.
The course teaches the Python code for Deep Learning step by step.
Forecasting the CO2 emissions all the way up to 2050!
All the steps needed in order to produce the most accurate forecasts possible!
The methodology can be applied to any time-series. Not only to CO2 emissions.

Forecasting CO2 emissions using Deep Learning

Requirements

There are no prerequisites because we build all the necessary knowledge slowly! Jump straight in!

Description

What is the course about:This online course shows how to apply Deep Learning in order to make forecasts far into the future e.g. year 2050. The focus is placed on CO2 emissions. Save time and learn step by step every line of code. Learn only the theory that is needed. No time to waste in abstract mathematics. This course is very applied and shows exactly how Deep Learning is implemented at workplace.We develop Deep Neural Network models and train them in an existing dataset of World Bank. Then these models are used to make forecasts all the way to 2050 ! This shows how impressive Deep Learning is!​There are no prerequisites. Every topic is analysed in depth so you can feel confident about what you learn.  This course is part of the official certificate for the ENERGY-COMPLETE dot com program for high-tech projects. Visit and download the latest version of the code!Who:Dr. Giannelos is a Research Scientist at Imperial College London, and has been part of high-tech projects at the intersection of Academia & Industry, prior to, during & after his Ph.D. Dr Giannelos is also the Principal Research Scientist for the ENERGY-COMPLETE dot com program. Join us!Doctor of Philosophy (Ph.D.) in Analytics & Mathematical Optimization applied to Energy Investments, from Imperial College London, and Masters of Engineering (M. Eng.) in Power Systems and Economics. Important:Prerequisites: Nothing is needed. Simply start learning by watching the videos. You will immediately start building confidence.Every detail is explained, so that you won’t have to search online, or guess. We start from scratch. You do not need to have done any preparatory work in advance at all.  Just follow what is shown on screen, because we go slowly and look at everything in detail.

Overview

Section 1: Intro

Lecture 1 Additions

Lecture 2 Summary

Lecture 3 General overview

Section 2: Preparing the data

Lecture 4 Data Preprocessing

Section 3: Polynomial Features

Lecture 5 Adding Polynomial Features in the models

Section 4: Dataset selection

Lecture 6 Splitting the data

Section 5: Scaling the data

Lecture 7 Scaling the features matrix and target variables

Section 6: Model definition & training

Lecture 8 Defining and compiling the Deep Neural Network model

Lecture 9 Fitting the Deep Neural network model

Lecture 10 Drawing the Deep Neural Network and clarifying the Activation Function

Lecture 11 Showing how to draw it in more detail

Section 7: Getting the predictions

Lecture 12 Generating the training and test set predictions

Section 8: Computing the test-MAPE

Lecture 13 Generating the test-errors

Section 9: Generating the train-set MAPE

Lecture 14 Training set Predictions

Section 10: Overfitting analysis

Lecture 15 Conducting overfitting analysis

Section 11: Sensitivity analysis

Lecture 16 Hyperparameter tuning versus Sensitivity Analysis

Lecture 17 Sensitivity analysis : Test set MAPE based on hyperparameters

Lecture 18 The naive model benchmark test

Section 12: Forecasts

Lecture 19 Preparing the forecasts: Theory, Methodology

Lecture 20 Generating the forecasts and plotting them

Lecture 21 Final Selection (filtering) of the models

Section 13: Bonus

Lecture 22 Some opportunities for you

Members of the ENERGY-COMPLETE dot com program,Enterpreneurs,Economists,Quants,Investment Bankers,Academics, PhD Students, MSc Students, Undergrads,Postgraduate and PhD students,Data Scientists,Energy professionals (investment planning, power system analysis),Software Engineers,Finance professionals

Course Information:

Udemy | English | 3h 45m | 1.74 GB
Created by: Dr. Spyros Giannelos

You Can See More Courses in the Developer >> Greetings from CourseDown.com

New Courses

Scroll to Top