辅导案例-IE5220

IE5220: Assignment
Time Series Prediction for weblink Traffic

The 1st Due Date: March 13rd, 2020 Friday at 23:59 EST
The 2nd Due Date: March 14rd, 2020 Saturday at 23:59 EST


Background
This assignment builds from our discussion of forecasting using different approaches to model
the pattern of website daily views and automatically select the best model to predict future daily
views for each website. Approaches are limited to three we discussed:
1. Time series clustering
2. Facebook Prophet
3. XGBoost

Data
File name: inputdata.csv
The input dataset was collected with a daily view over a period between Jan 2017 and Jan 2020
(37 months) were measured. It consists of approximately 1,093 time series. Each of these time
series represents a number of daily views of a different weblink, starting from January 1st, 2016
up until January 31st, 2020. The goal is to forecast the daily views into the future for each
weblink in the dataset from February 1st, 2020 to December 31st, 2020.

Requirements
1. Use Python 3.6 or 3.7
2. Turn in your code and figures along with summary based on your observation of model
performance





51作业君 51作业君

Email:51zuoyejun

@gmail.com

添加客服微信: ITCSdaixie