Forecasting ATM Dispense

Hi Friends,

I have a challenge in forecasting the daily dispense of a set of ATMs. I appreciate if some one can help me in it. Requirements listed below:

  1. Forecasting for Multiple sites. How can I do it?
  2. How to give a forecast for new set of ATMs which have been recently installed?
  3. Dispense varies during weekends, Month Beginning and End, Seasonal etc.,
  4. when an ATM goes cashout, obviously its dispense will drop and that data point becomes an outlier in the dataset. Any suitable technique we can use for this?

It would be great if some one can help me on this.

Thanks in advance

Best regards,
Neel

1 Like

HI Neel,

we need to Explore the data distributions for ATMs data to improve the operational efficiency of ATMs

No. of transactions per day
Amount withdrawn per day
No. of Customers using an ATM
Identify potential Customers for cross selling
Amount of transactions by each Customer
Identify areas to launch new products or marketing campaign.
Amount of transactions in an area
and you can use multiple time series for forecasting for different sites
you can drop outlier and proceed with evaluating the assumption of time series and then build a forecasting model

Multiple Time Series

We can plot multiple time series in one chart by combining both the series into a matrix.

Get the data points in form of a R vector.

rainfall1 <- c(799,1174.8,865.1,1334.6,635.4,918.5,685.5,998.6,784.2,985,882.8,1071)
rainfall2 <- c(655,1306.9,1323.4,1172.2,562.2,824,822.4,1265.5,799.6,1105.6,1106.7,1337.8)

Convert them to a matrix.

combined.rainfall <- matrix(c(rainfall1,rainfall2),nrow = 12)

Convert it to a time series object.

rainfall.timeseries <- ts(combined.rainfall,start = c(2012,1),frequency = 12)

Print the timeseries data.

print(rainfall.timeseries)

Plot a graph of the time series.

plot(rainfall.timeseries, main = “Multiple Time Series”)

Regards,
Tony

© Copyright 2013-2019 Analytics Vidhya