CPS Power BI – Tensorflow Regressor

“CPS Power BI – Tensorflow Regressor” is a powerful Machine Learning tool tha enables single/multiple linear/nonlinear regression interactive modeling within Microsoft Power BI without requiring users to write a single line a code.
Users can also save models in order to use them across different report visuals and production environment applications.
Check CPS PBI TF Regressor at work in the following Power BI report, add Layers by simply clicking the “ADD LAYER” button and specify each layer’s neurons to interactively configure a Neural Network, then launch the training and check model convergence as it happens.
Adjust the total number of iterations if needed for more complex curves and inspect each Model Topology by saving it on your device.

Yes we did it!

We’ve just successfully tested an almost seamless integration of the most popular  Machine Learning framework, Tensorflow, into the most intuitive and yet powerful business analytics tool, Microsoft Power BI by developing an extremely easy to use Power BI Custom Visual thus opening to unlimited possibilities in terms of advanced zero coding data analysis and modeling in Power BI.

TensorFlow is a software library for machine learning and artificial intelligence developed and mantained by Google. It can be used across a range of tasks, but is used mainly for training and inference of neural networks. Tensorflow is one of the most popular high performance deep learning frameworks, alongside with PyTorch.

Our first experiment consists of a regression engine or “modeler” for the identification of descriptive and predictive models though multiple linear and non-linear regressions. Whatch the demo video here:

In this demo, the Tensorflow Regressor Custom Visual asynchronously trains three different models:

1. A simple Multiple Linear Regression Model made of 1 Layer and 1 Neuron;
2. A Neural Network Model made of 2 Layers (1 hidden with 100 Neurons);
3. A Neural Network Model made of 3 Layers (2 hidden with 50 Neuron each).

A similar interactive demo is available online at: DATITALIA – CPS  at report page 7.

The next step will be the development of a Power Bi Custom Visual that allows the interactive configuration of a custom neural network for multiple regression modeling with Tensorflow in Power BI as well as model saving and reloading without writing a single line of code.

After having set up the kind of model, users will be able to experiment each model by simply dragging Power BI columns in the features and outcome fields on the Visualizations Panel, adjust the number of epochs and save the model for further analysis on incoming data.

Refer to analytics@cpsweb.it for more details and requests.

CPS Correlogram++ for Power BI adds colored/custom labeled interactive scatterplot analysis to further speed-up Data Exploration by condensing even more key preliminary information about correlations and distributions in the same intuitive and interactive graphic contex.

(nov. 2024)

Users can now select a column for labeling datapoints and then select label values by simply clicking an item in the legend to have the corresponding points highlighted in the scatterplot thus further facilitating co-variates selection and filtering.

The scatterplot header reports the formula of the regression line.

…climbed another level up…

Watch the full video tutorial…

CPS Correlogram+ for Power BI is a custom Power BI Visual thought to facilitate Data Exploration by condensing key preliminary information about correlations and distributions in the same intuitive and interactive graphic contex.

(Nov. 2024)

LET’S SEE HOW IT WORKS

Correlation tiles – we started by emulating the most popular tool for plotting Correlograms that’s undoubtedly Cran R’s “corrplot” so, by simply dragging numeric columns in the only field, named “Column”, in the Power Visualizations panel (filtering out null/blank values) you’ll see a tile matrix in which tiles different colors and size depend on the Pearson’s correlation coefficient between the values of the two columns whose names are on the row and column of the matrix that cross with in correspondence with each tile.

CPS Corrplot+ supports up to 40’000 rows and 250’000 values. Such amounts might be increased in case of customers’ special needs.

When the correlogram appears a text that will fade-out in few seconds notifies the user about how many rows and values are involved in the calculation of correlation coefficients as this is also useful to know the exact size of the sample especially in case of further direct or indirect filterings.

Once the user click on a tile a kind of custom tooltip will show the scatterplot of the distribution of points whose coordinates are: the values of column the row of the tile for the X axis and those on the column of the tile for the Y axis.

Scatterplot’s header shows the equation of the regression line that will be also displayed on the scatter plot.

In order to ensure scatterplot readability the X and the Y axis have independent scales avoiding that, for example, in the case a percentage (ranging from 1 to 100) on the X axis and an average incomes (or sales amounts) on the Y axis (or vice versa), percentages get plotted extremely close to the other axis. Please note that a relevant difference between the X and the Y scale will affect the correspondence between the slope of the regression line on the scatterplot and it’s angular coefficient reported on the equation on the header of the scatterplot but will not affect the relation with points.

You can check CPS Correlogram in action at on this link DATITALIA – CPS on the report pages concerning correlation analysis or refer to the demo video.

Do not hesitate to mail to: analytics@cpsweb.it for requests or further info.

CPS CUSTOM VISUALS FOR MICROSOFT POWER BI

Refer to analytics@cpsweb.it for requests and more details.

CPS PROVIDES WORLD CLASS STRATEGY AND FUNCTIONAL CONSULTING SERVICES IN THE FOLLOWING FIELDS:

Marketing & Communication 

Supply Chain, Logistics & Operation

Information Technology 

Procurement and Tenders

WE ALSO PROVIDE THE FOLLOWING CROSS-FUNCTIONAL SERVICES

Strategy Execution/Implementation Management

Business Analytics

Program/Project Management

DATA-DRIVEN CREATIVITY

I Business Analytics sono la più importante risorsa tecnologica a supporto del disegno e dell’esecuzione delle strategie dell’azienda moderna.
Nella grande maggioranza dei casi le aziende sono ferme alla visualizzazione di dati aggregati in dashboard quasi sempre troppo prolisse e poco utili, scarsamente analitiche nonostante la definizione dei relativi KPI abbia richiesto mesi.
Invece, il potenziale di ciò che i produttori di software chiamano “Advanced Analytics” eventualmente associati a “Machine Learning” e “Artifical Intelligence” può essere enorme sia in termini di sviluppo che di razionalizzazione del business.
In molti casi mancano le competenze per dare adeguato seguito alle intuizioni o alle richieste del top-management.
Il fatto è che per trarre reali vantaggi dai Business Analytics servono competenze sofisticate, riferite a nuove figure professionali e… “l’esperienza non guasta”.

COMPETENZA
I Business Analytics rappresentano la più importante competenza aziendale a supporto dello sviluppo e dell’esecuzione strategica in quanto in sono grado di abilitare in modo rapido e solido quel processo di Discovery tipico dell’innovazione attraverso la previsione e la valutazione di scenari, il test delle ipotesi e la simulazione di modelli alternativi, suggerire o prescrivere adeguamenti delle strategie, delle tattiche o dei processi.
Troppe aziende tendono a riferire i Business Analytics a risorse IT, non istituiscono uno stretto legame tra queste e il business e confidano in modo sconsiderato sugli strumenti che, invece, altro non sono che enabler più o meno inutili in assenza di metodo e capacità analitiche.
CPS stravolge da sempre simili approcci partendo sempre dall’analisi strategica e dalla misura di quello che veramente conta, fornendo competenze realmente integrate dalla strategia all’esecuzione in un processo creativo ma rigoroso guidato dall’evidenza dei dati.

L’applicazione degli Analytics al marketing (non solo digitale), alle operazioni e ai processi o alle varie fasi del Service Delivery Lifecycle e alla gestione degli asset tecnologici si traduce in una maggiore comprensione del contesto, dei fenomeni che lo caratterizano e delle variabili che lo influenzano, in una più profonda comprensione della propria utenza, nella capacità di classificare le priorità e quindi di misurare l’impatto degli investimenti o di semplici cambiamenti in modo solido e adattativo.

ESPERIENZA
Grazie ad una notevole serie di esperienze di successo CPS è in grado di associare competenze certificate di massimo livello nel campo della Pianificazione e dell’esecuzione strategica e nel campo della strutturazione e dell’analisi dei dati senza trascurare i problemi connessi alla sicurezza degli stessi e alla compliance normativa del loro trattemento.

METODO
CPS ha sviluppato, e continua a perfezionare, metodologie di approccio ai business analytics specifiche in funzione dei settori di applicazione quali: Marketing & Communication, Supply Chain & Logistics, Operation e IT.

Se pensate che non sia più sufficente dotarsi di strumenti potenti o che ci sia qualcosa di importante oltre più o meno sofisticate visualizzazioni di dati aggregati, siete sulla strada giusta.
…se avete bisogno di un partner solido e concreto anche solo per validare il vostro approccio metteteci alla prova.

Privacy Policy