openSAP Become and Augmented BI Expert. Week 4 Augmented Analytics

Hello, we continue in week 4, in which if you want to know everything we can do with Search Insight, Smart Discovery and predictive scenarios such as Forecasting , Classification and Regression, this is your week !!!.

I think it is a very interesting week to be able to practice with predictive scenarios, which can serve as an example for the projects we are doing. We must remember that we can apply it to our planning processes.

Continue?

Search to insight

We start with the option of Search to Insight, search for results with natural language.  In this video they explain what options the user has to be able to ask SAC for the data and how SAC returns this data. For this we have to follow some conditions and restrictions to reach the result that the user needs. It’s NOT just asking…

This option works with both purchased models and Live, BW, HANA, S/4 HANA models.  Not in all models we can ask but those that are indexed and the most important thing is that it respects the authorizations both those of SAC and those of the Live connections.  In the video they give us a series of recommendations that we must follow.  Smart to Insight is an interesting option for the end user. Advisable.

Improve your data understanding with in-story augmented insights

We continue with Smart Insights within our stories and how we can get the «hidden» data, who are the most relevant contributors, The why sales are higher?, how the data changes. They explain how we can configure Smart Insights, to exclude dimensions or even values of a dimension. Realize more dynamic Insights.

Enable Smart Insights for SAP HANA Live connection, enable automatic forecasting in time series charts, line charts, or schedule tables with the click of a button.

Automated data exploration with smart discovery

With this video we will understand and how to use, Smart Discovery.  They explain that Smart Discovery is the automated data exploration capability that allows our user to use machine learning to answer a business question.  The good thing about this is that it can be applied directly to BI data and does not require any external data preparation. Smart Discovery can be run from within a story or directly from the homepage when creating a new story. It all starts with defining the business question we want to answer. The business question is described as a destination, which is the dimension or indicator in the data we want to understand, and an entity, which is the object in the data around which we want to perform our analysis.

Once the question is asked, a series of pages appear, the first is a summary page that offers a high-level view of the data relevant to your question. The second page shown is the key  influencing factors. The third of those values that are unexpected and a last page of simulation allows the user to interact with the influencing factors directly. It really is interesting and how we can give more autonomy to the end user.

Prepare the right data for predictive scenarios

In the following video they explain how we can use predictive scenarios, the idea of predictive scenarios is to bring the power of machine learning automation to planners and end users.  Without having to have the intervention of a data scientist. Depending on the data, we have different scenarios, if they are data, revenue or cost forecast, we will use time series, business results values, if they use regression or if we want to anticipate the behavior of the client or employee, they can take advantage of the classification.

The options we have where to apply these scenarios are three, planning, acquired data and for data sets with Live connection.

In this video they help us to start the preparation of the data and with several examples that take away «the fear» of using predictions. Very well explained.

Predict future outcomes with Smart Predict Forecasting

Once they have explained each of the scenarios, they start with Smart Predict Forecasting. Where they put more emphasis is to use it in planinng, since Classidication and Regression is NOT possible to use it in planning.

They explain step by step how to configure the scenario and how to interpret the data.

It is one of the scenarios that we can apply in our projects in a very easy way. I recommend you watch the video is very interesting.

Predict event probabilities with Smart Predict Classification

In this video they explain how to use the Classification scenario, how to configure it and get to apply a Classification scenario. As they indicate we have to start with:

What is it?
A classification model helps us classify observations based on historical data.
Using classification, the probability of a specific event occurring can be predicted.
What use cases?
Questions can be addressed through classification.
Who is likely to be based on ?
Examples:
Who is likely to win a horse race based on its recent form?
Who is likely to buy this product based on their buying habits?
Who is likely to leave a company based on their employee profile?
Who is likely to stop using a service based on its use and satisfaction?

Predict key indicators with Smart Predict Regression

We finish with the explanation of the Regression scenario and as in the previous one they help us to understand it and how to apply it:

What is it?
A regression model is used to estimate the value of a measure.
Using a regression model, the most likely value for a given observation can be predicted.
What use cases ?
The questions that can be addressed by regression.
What is the ?
How many/how much?
Examples:
What is the delay of each customer in paying their invoices?
How many products will a customer buy in the next quarter?
How much will a customer spend on my e-commerce on average?

In short, it is a very interesting week to be able to apply in any project, BI or Planning and that gives more value to SAC than its competitors may have. Very well explained.

I hope it serves them

Continuing to move forward, knowing is good and also, knowing who knows, can help us 

Publicado por Óscar Gómez

Os dejo mi perfil https://www.linkedin.com/in/oscar-g%C3%B3mez-4b78a62a/

Deja una respuesta

Introduce tus datos o haz clic en un icono para iniciar sesión:

Logo de WordPress.com

Estás comentando usando tu cuenta de WordPress.com. Salir /  Cambiar )

Imagen de Twitter

Estás comentando usando tu cuenta de Twitter. Salir /  Cambiar )

Foto de Facebook

Estás comentando usando tu cuenta de Facebook. Salir /  Cambiar )

Conectando a %s

A %d blogueros les gusta esto: