Absolutely, Machine Learning can be very useful in the situation you have described for planning your Winter 25 purchases.
You need to combine the data you have for the past years (historical data), clean and process it so that it is formatted correctly for Qlik AutoML. Define the target variable (sales) and the features. The historical sales data will be considered as input features to predict the outcome.
Once you create the AutoML experiment, the AutoML will automatically test the data with different ML algorithms and identify the best model for prediction.
Please follow the attached help documents and create a test experiment:
here is un example data . Our product has 6 featurs and sum of quantity and sales of previous years. as it described. we need to purchase products with the same featurs for next year. Our goal is maximize the sales. And to purchase products with these features that will sold out , but to keep enough inventory for the all year.
so the goal is to purchase the exact amount of products for each features.
(we can purcahse 1 time a year). so a-f are featres and the quantity and sales are time features so how do I put them for qlik as time features and not just as a number??
From the sample excel file, it seems like you have data with features (a-f) coupled with historical sales and quantity information for previous years. Although you cannot input features (a-f) as time features, you can load the data to Auto ML, structure your data to capture the historical aspect and use it for predictions. Please note that most ML algorithms require variables to be numeric. For categorical feature handling, Auto ML uses the concept of one hot encoding automatically.
I would suggest you to add additional fields for your data model such as Sales growth rate YoY to estimate the demand. Please have a look at the attached video which shows a complete basic walk-through of Qlik AutoML and creating a simple predictive analytics application in Qlik Sense SaaS Enterprise.
@dtbit123 You don't have to manually color code, that will be taken care of by one hot encoding. The Qlik Auto ML does it for you by default.
Just to give a simple example of how one hot encoding works in ML, please review the following example:
Let's say I have three fruits Apple, Orange, and Banana, essentially one hot encoding converts the categorical features into a vector-based value so that the machines can understand.
Apple Orange Banana 1 0 0 0 1 0 0 0 1
Here, 1 marks the presence of that category.
I would request you to follow the video and create a test ML experiment by following the sample example so that you can explore and understand the product.
