Description

Data
The goal of this homework is to develop a method to predict the electricity usage based on the weather conditions. We provide the following two datasets for this task:
1. Weather: Weather data for one year with daily weather conditions
2. Energy Usage: Energy usage history for one year (in kW) with 30-minute intervals. The energy usage of specific devices like AC, Fridge, washer, etc. are also given.
Tasks
1. Examine the data, parse the time fields wherever necessary. Take the sum of the energy usage (Use [kW]) to get per day usage and merge it with weather data (10 Points).
3. Linear Regression – Predicting Energy Usage:
4. Logistic Regression – Temperature classification:
Finally generate a csv dump of the classification (1 for high, 0 for low)
(20 points)
5. Energy usage data Analysis:
We want to analyze how different devices are being used in different times of the day.
– Is the washer being used only during the day?
– During what time of the day is AC used most?
There are a number of questions that can be asked.
For simplicity, let’s divide a day in two parts:
– Day: 6AM – 7PM
– Night: 7PM – 6AM
Analyze the usage of any two devices of your choice during the ‘day’ and ‘night’. Plot these trends. Explain your findings. (10 points)
6. Visual Appeal and Layout – For all the tasks above, please include an explanation wherever asked and make sure that your procedure is documented (suitable comments) as good as you can.
Don’t forget to label all plots and include legends wherever necessary as this is key to making good visualizations! Ensure that the plots are visible enough by playing with size parameters. Be sure to use appropriate color schemes wherever possible to maximize the ease of understandability. Everything must be laid out in a python notebook (.ipynb). (5 Point)
Submission
2. If you do not have much experience with Python and the associated tools, this homework will be a substantial amount of work. Get started on it as early as possible!
3. Please use Piazza to ask any questions.
4. Submit everything through Blackboard. You will need to upload:
a. The Jupyter notebook all your work is in (.ipynb file)
b. Python file (export the notebook as .py)
c. PDF (export the notebook as a pdf file)
d. Linear regression and logistic regression csv dumps
These files should be named with the following format, where the italicized parts should be replaced with the corresponding values:
1. cse351_hw2_lastname_firstname_sbuid.ipynb
2. cse351_hw2_lastname_firstname_sbuid.py
3. cse351_hw2_lastname_firstname_sbuid.pdf
4. cse351_hw2_lastname_firstname_sbuid_linear_regression.csv
5. cse351_hw2_lastname_firstname_sbuid_logistic_regression.csv