parthakuila/Time-Series-Analysis-and-Forecasting parthakuila/Time-Series-Analysis-and-Forecasting

  • Stars
    star
    2
  • Language
    R
  • Created almost 6 years ago
  • Updated over 4 years ago
Twitter logo Share LinkedIn logo Share Facebook logo Share
parthakuila/Time-Series-Analysis-and-Forecasting
github

parthakuila

Reviews

There are no reviews yet. Be the first to send feedback to the community and the maintainers!

Repository Details

Build time series modelling and evaluation the best model. Build model like ARIMA, Exponential Smoothing, TBATS, NN and others. Compare the best model forecast the next 30 , 60 data points.

More Repositories

1

Customer-Lifetime-Value-Prediction

Using data: Customer's invoice file. Introductions: Customer Lifetime Value(CLTV) "Customer Lifetime Value is a monetary value that represents the amount of revenue or profit a customer will give the company over the period of the relationship". CLTV demonstrates the implications of acquiring long-term customers compare to short-term customers. Customer lifetime value (CLV) can help you to answers the most important questions about sales to every company: How to Identify the most profitable customers? How can a company offer the best product and make the most money? How to segment profitable customers? How much budget need to spend to acquire customers? CLTV indicates the total revenue from the customer during the entire relationship. CLTV helps companies to focus on those potential customers who can bring in more revenue in the future. CLTV = ((Average Order Value x Purchase Frequency)/Churn Rate) x Profit margin. Please check the below step for how to calculate CLTV. Algorithm: Step1: Calculate CLTV. Calculate the average order value of customers: Average order value = Total money spent / total number of transactions Calculate Purchase Frequency: Purchase Frequency = Total Number of Orders / Total Number of Customers Calculate Repeat rate and Churn rate: Repeat rate = How many customers have numbers of transactions more than one / total numbers of customers Churn rate = 1 - repeat rate Calculate the profit margin: Profit margin is the commonly used profitability ratio. It represents how much percentage of total sales has earned as the gain. Let's assume our business has approx 5% profit on the total sale. Profit margin = Total money spent on each customer * 0.05 Calculate customer lifetime value: Customer value = (Average Order Value * Purchase Frequency) / Churn rate Customer lifetime value = Customer value * Profit margin Step2: Predictive modelling. Build a regression model for existing customers. Take recent six-month data as independent variables and total revenue over existing time( here taking 2 years) as a dependent variable and build a regression model on this data. Pros and Cons of CLTV: CLTV helps you to design an effective business plan and also provide a chance to scale your business. CLTV draw meaningful customer segments these segment can help you to identify the needs of the different-different segment. Customer Lifetime Value is a tool, not a strategy. CLTV can figure out the most profitable customers, but how you are going to make a profit from them, it depends on your strategy. Generally, CLTV models are confused and misused. Obsession with CLTV may create blinders. Companies only focus on finding the best customer group and focusing on them and repeat the business, but it’s also important to give attention to other customers.
Python
6
star
2

Predicting-a-Player-s-position-based-on-the-attributes

Predicting a Player’s position based on the attributes Using the data, create a model that accurately predicts/assigns a players position based on the individual attributes. The following steps should be clearly elucidated: 1. Data Cleaning 2. Features considered for EDA and further steps. 3. Exploratory Data Analysis Undertaken 4. Inference from EDA 5. Choice of Best Algorithm and Why 6. Training Accuracy 7. Predictions with test data
Jupyter Notebook
2
star
3

Text-topic-modelling

Predict if a particular sentence in an article should be included in the summary of the article or not (Incorporates NLP Along with ML)
R
1
star
4

Twitter-Data-Analysis

Web Scraping tweets from my personal api. Using R Language for web Scraping and analysis and Visualization.
R
1
star
5

Network-fault-incident-prediction

https://docs.google.com/document/d/1gyzClRjjgGy2lyoqyhqAO1amBIEEDmbXRx_-tqmdMDs/view
Python
1
star
6

Time_Series-Analysis-Using-Python

olt(Optical Line Termination) forecasting, build time series modelling for forecasting.
Python
1
star
7

Churn-Analysis-and-Prediction-Modelling

Customer Behavior, Feature Selection, churn prediction. Develop ML model for churn Prediction. Used ML model Like LR, LDA, KNN, CART, NB, RF, XGB, AdaBoost, SVM.
Python
1
star
8

Sentiment-analysis-for-Amazon-reviews-data

R
1
star
9

Python-Script-for-dump-excel-file-within-Database.

Dump excel file within Database. Use MongoDb database. Site wise collection dump into the database. Use library pymongo, pandas, datetime, jason, time and others. At first connect with database and load the data. Then dump the data specific site wise collection.
Python
1
star
10

Ernst-Young-Case-study-of-ML

Problem: Predict the sales demand for consumer goods. Data: Attached is a spreadsheet containing sales data. The attached document contains instruction and clarification about the data. Please follow the instructions and prepare the output file. Case Study will be evaluated on the below criteria Data Processing Feature Engineering Code Automation
R
1
star
11

Predicting-Future-Sales

Provided with daily historical sales data. The task is to forecast the total amount of products sold in every shop for the test set. Note that the list of shops and products slightly changes every month. Creating a robust model that can handle such situations is part of the challenge.
Jupyter Notebook
1
star
12

Recharge-Use-Cases-Analysis

Subscriber wise daily usage analysis and Bucketing daily typical use wise. Clustering several group of Bucket for recommendation Daily typical usage per subscriber over telecom network.
Python
1
star