[go: up one dir, main page]
Scribd
Upload
12 views2 pages

Assignment Entry Level

The assignment focuses on time series analysis and forecasting using a dataset of weekly advertising spends, discounts, and sales across brands. Key tasks include data preprocessing, outlier detection, exploratory data analysis, identifying non-forecastable series, and building a forecasting model. Submissions should include well-commented code, visualizations, justifications for methods used, and a regression model with performance metrics.

Uploaded by

Recovery Account
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
12 views2 pages

Assignment Entry Level

The assignment focuses on time series analysis and forecasting using a dataset of weekly advertising spends, discounts, and sales across brands. Key tasks include data preprocessing, outlier detection, exploratory data analysis, identifying non-forecastable series, and building a forecasting model. Submissions should include well-commented code, visualizations, justifications for methods used, and a regression model with performance metrics.

Uploaded by

Recovery Account
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 2

Data Science Assignment – Time Series Analysis and Forecasting

Duration: 30 minutes

Difficulty Level: Entry-Level / Freshmen

Objective
You are provided with a dataset containing weekly advertising spends, discount percentages, unit
sales, and total sales values across multiple brands and geographic regions. Your task is to process
the dataset, handle missing values, identify anomalies, conduct exploratory data analysis (EDA),
and build a forecasting model.

Assignment Tasks
1. Data Preprocessing (5 minutes)
• Load the dataset.
• Check for missing dates and fill them appropriately.
• Impute missing values for the numerical columns using a suitable method.

2. Outlier Detection (5 minutes)


• Identify 2-3 outliers present in the dataset.
• Explain the method used to detect outliers.
• Suggest a method to handle them.

3. Exploratory Data Analysis (EDA) (7 minutes)


• Expected:
• Time series trend analysis of total sales value.
• Correlation between media spends and sales.
• Good to have:
• Visualization of discount trends across brands.
• Analysis of the impact of discount on unit sales.

4. Identify Non-Forecastable Series (5 minutes)


• Identify any non-forecastable time series (if present) based on patterns like randomness or
lack of trend.
• Justify why certain series cannot be forecasted effectively.

5. Forecasting Model (8 minutes)


• Expected:
• Build a simple regression model to forecast two months of sales using media
spends, discounts, and any other relevant features.
• Good to have:
• Evaluate model performance using RMSE or MAE.
• Compare a simple regression model with a time-series model like ARIMA or
XGBoost (optional).

What’s Expected in Submission?


✅ Well-commented code explaining each step.
✅ Basic visualizations to support findings.
✅ Justifications for methods used in missing value imputation, outlier detection, and non-
forecastable series identification.
✅ A simple regression model with at least one performance metric.

Bonus (Optional, If Time Permits)


🌟 Experiment with feature engineering to improve the model.
🌟 Identify seasonality patterns in sales data.
🌟 Compare different forecasting models.

This assignment is designed to assess data wrangling, EDA, anomaly detection, forecasting, and
problem-solving skills.
Good luck! 🚀

-----------------------------------------------------------------------
Data
[Find data here]

You might also like