# Predicting electricity consumption in a household

Please I need your help in brainstorming on a problem. I have a community, in this community, every household has a 'power device' that powers the household. Households need to buy power units from a power company, and then, the 'power device' is loaded with the units in order to power the household. For example, if a customer pays 1000, he will be given units worth 9.3 to use in powering his household. This is sample of the data I have -

Customer ID | Date     | Amount | Units
3030         2019/01/23   1000    9.3
3829         2019/04/01   500     4.3
3030         2019/01/31   1000    9.3


My challenge -

Given this dataset, I want to be able to predict the units remaining in the 'power device' at any point in time. For example, looking at the dataset above, how many units will be remaining as at 2019/04/04 for customer ID - 3829?

Is this dataset sufficient and if yes, what ML algorithm would you suggest?

• It seems to me that you need some information about the rate of power consumption. Do you have such information? Jul 14, 2019 at 12:37
• No I don't have that. Jul 14, 2019 at 21:42
• Then the dataset is not sufficient. Jul 14, 2019 at 21:52
• Ok then. Thank you. Jul 14, 2019 at 21:54