Forecasting Analytics Group members: - Arpita - Kapil - - - PowerPoint PPT Presentation

▶
forecasting analytics
SMART_READER_LITE
LIVE PREVIEW

Forecasting Analytics Group members: - Arpita - Kapil - - - PowerPoint PPT Presentation

Mar 11, 2024 260 likes •462 views

Forecasting Analytics Group members: - Arpita - Kapil - Kaushik - Ridhima - Ushhan Business Problem Forecast daily sales of dairy products (excluding milk) to make a good prediction of future demand, and predict the stock level

slide-1
SLIDE 1

Forecasting Analytics

Group members:

  • Arpita
  • Kapil
  • Kaushik
  • Ridhima
  • Ushhan
slide-2
SLIDE 2

Marketing and Communications Council Class of 2013

Business Problem

  • Forecast daily sales of dairy products (excluding milk) to make a

good prediction of future demand, and predict the stock level required to meet the demand.

  • Evaluate different forecasting methods on data distribution and

forecast period, and pick the best one based on the results.

2

Forecasting Analysis Dairy Products Lassi

Srikhand

Ice Cream Cups & Cones Family Packs Family Packs Family Packs

slide-3
SLIDE 3

Marketing and Communications Council Class of 2013

Ice Cream & Gelato – Analysis & Forecast

3

slide-4
SLIDE 4

Marketing and Communications Council Class of 2013

4

2 days before Muharram Valentine’s day 1 days before Mahavir Jayanti Good Friday 31st March A day before Raksha Bnadhan Budha Purnima Gandhi Jayanti

Ice Cream Sales (2011 – 12)

Forecasting Analysis

slide-5
SLIDE 5

Marketing and Communications Council Class of 2013

Data Visualization- Sales

5

Forecasting Analysis

Distribution Across subclasses Week day seasonality Family Pack - Monthly Cups & Cones - Monthly

slide-6
SLIDE 6

Marketing and Communications Council Class of 2013

Analysis & Forecast – Sales ( Cups and Cones )

6

Forecasting Analysis

Regression ( Single Seasonality) Moving Average Method Holts Winter Method

MAPE 0.89 RMSE 13.22 MAPE 0.58 RMSE 16.46 MAPE 0.69 RMSE 8.75

slide-7
SLIDE 7

Marketing and Communications Council Class of 2013

Analysis & Forecast – Sales ( Cups and Cones - Continued)

7

Forecasting Analysis

Regression – Polynomial Trend and Multiple Seasonality ( Weekly and Half Yearly)

The Regression Model

Input variables Coefficient

  • Std. Error

p-value SS Constant term 11.43823338 2.29202032 0.00000094 87293.44531 t 0.07057966 0.02275607 0.0020786 1976.065308 t ^2

  • 0.00017919

0.00006641 0.00730373 43.97101974 day of week_2

  • 5.56998158

2.0351541 0.00651356 139.6604157 day of week_3

  • 5.55328846

2.03511047 0.00667317 196.7801056 day of week_4

  • 4.37169743

2.04480267 0.03320085 51.11870575 day of week_5

  • 6.26070452

2.0447247 0.00236676 629.9749756 day of week_6

  • 3.29685378

2.0446682 0.10775778 89.6808548 day of week_7

  • 3.32565784

2.04463482 0.10472196 284.4305725 month of year_1 6.26655626 1.61968482 0.00012995 1627.033813

Training Data scoring - Summary Report

Total sum of squared errors RMS Error Average Error 38694.54028 10.28215871

  • 0.0001072

Validation Data scoring - Summary Report

Total sum of squared errors RMS Error Average Error MAPE 5471.241832 13.28501682 3.254043135 0.55182345

MAPE improved

slide-8
SLIDE 8

Marketing and Communications Council Class of 2013

Lassi & Srikhand

8

slide-9
SLIDE 9

Marketing and Communications Council Class of 2013

Data Visualization-Weekly Demand

9

Forecasting Analysis

slide-10
SLIDE 10

Marketing and Communications Council Class of 2013

Daily Quantity Sold Data

10

Forecasting Analysis

slide-11
SLIDE 11

Marketing and Communications Council Class of 2013

Correlation between daily sales

11

  • 1
  • 0.5

0.5 1 1 2 3 4 5 6 7 ACF Lags

ACF Plot for Sum of Quantity_Sold

ACF UCI LCI

Next day related to the previous day and a week before Forecasting Analysis

slide-12
SLIDE 12

Marketing and Communications Council Class of 2013

Naïve Forecast

  • Negative correlation between a day's sales and sales previous day? Using Naïve

we are just taking previous day and forecasting but it seems lot of signal is not captured

  • It says that naïve forecasting model is not able to explain

12

  • 1
  • 0.5

0.5 1 1 2 3 4 5 6 7 ACF Lags

ACF Plot for Residual

ACF UCI LCI

MAPE 130% Forecasting Analysis

slide-13
SLIDE 13

Marketing and Communications Council Class of 2013

Forecast using Multiple Regression

Input variables Coefficient Std. Error p-value SS Constant term

  • 1936.91626 406.079040

5 0.00000276 203580.109 4

Row Labels

0.04776152 0.00991758 0.00000222 7851.68261 7

Weekday_Mon

  • 2.3526628 3.8025558 0.53654903 4072.53735

4

Weekday_Sat

9.82291985 3.802526 0.01022904 342.777923 6

Weekday_Sun

26.75549126 3.82146454 0 24514.2578 1

Weekday_Thu

4.24133205 3.88327575 0.27556217 456.080169 7

Weekday_Tue

  • 2.26675916 3.92769527 0.56426156 535.565490

7

Weekday_Wed

4.0734067 3.92513132 0.30015546 373.083984 4

13

Residual df 321 Multiple R-squared 0.255420386

  • Std. Dev. estimate

18.61228561 Residual SS 111199.9063

Training Data scoring - Summary Report

Total sum of squared errors RMS Error Average Error 111199.9052 18.38460315

  • 6.71529E-06

Validation Data scoring - Summary Report

Total sum of squared errors RMS Error Average Error 9126.699565 18.05418389

  • 11.31025751

MAPE: 133% Forecasting Analysis

slide-14
SLIDE 14

Marketing and Communications Council Class of 2013

ACF & PACF with Multiple Regression

  • Regression model tells us that next day’s

sales are dependent on last two days’ sales along with regular time component

14

  • 1
  • 0.5

0.5 1 1 2 3 4 5 6 7 ACF Lags

ACF Plot for Residual

ACF UCI LCI

  • 1
  • 0.5

0.5 1 1 2 3 4 5 6 7 PACF Lags

PACF Plot for Residual

PACF UCI LCI

Forecasting Analysis

slide-15
SLIDE 15

Marketing and Communications Council Class of 2013

Smoothing Moving Average(7)

Error Measures (Training)

MAPE 128.55637 MAD 17.43578 MSE 605.26376

15

20 40 60 80 100 120 140 160 Sum of Quantity_Sold Row Labels

Time Plot of Actual Vs Forecast (Training Data)

Actual Forecast 10 20 30 40 50 60 70 Sum of Quantity_Sold Row Labels

Time Plot of Actual Vs Forecast (Validation Data)

Actual Forecast

Error Measures (Validation)

MAPE 241.73763 MAD 27.321429 MSE 866.89286

Shows strong correlation between days’ sales with last 2 days It also tells that moving average could be used in roll forward manner Forecasting Analysis

slide-16
SLIDE 16

Marketing and Communications Council Class of 2013

Holt Winter Forecasting Method-Additive

16 Error Measures (Validation)

MAPE 56.46093 MAD 18.750176 MSE 367.86672

Error Measures (Training)

MAPE 121.54953 MAD 15.887538 MSE 494.58883

Forecasting Analysis

slide-17
SLIDE 17

Marketing and Communications Council Class of 2013

Moving Average (MA2)

  • Model Fits quite well but can only forecast

for next 1-2 days

17

MAPE 123.82985 MAD 16.119131 MSE 513.18213 MAPE 48.290598 MAD 16.5 MSE 308.25

Forecasting Analysis

slide-18
SLIDE 18

Marketing and Communications Council Class of 2013

Holt Winter Smoothing-Additive

18 Error Measures (Training)

MAPE 125.72335 MAD 17.082153 MSE 577.27479

Error Measures (Validation)

MAPE 48.290598 MAD 16.5 MSE 308.25

Forecasting Analysis

slide-19
SLIDE 19

Marketing and Communications Council Class of 2013

Conclusions & Suggestions

  • Different in forecasting methods used for effective

forecast

  • Recommendations to Business
  • Stock level on weekly or monthly basis can be

predicted for dairy products.

  • ERP system that could directly tell the vendor how

much to deliver

  • Model is useful to predict for next day given previous

two days sales, need to implement roll forward

19

Forecasting Analysis

slide-20
SLIDE 20

Marketing and Communications Council Class of 2013

THANK YOU

20