topic Possible Machine Line learning types we can perform in Qlik Auto ML? in Qlik Predict

Possible Machine Line learning types we can perform in Qlik Auto ML?

Venkadesh_Ponnu — Thu, 01 Sep 2022 11:39:53 GMT

Here I am using Qlik Auto ML to create and deploy ML model.

In demo/Training videos we are using Supervised Learning i.e. passing the historical data and train the model. then we will predict the target column mostly yes/no type. (Categorial)

Now my question 1 is - what are the Machine learning types we can use it in Qlik Auto ML other than Supervised Categorial type? (please answer as per features available on 1st sept 2022)

I am trying to predict the stock price i.e. Close Price. I am unable to complete it due to insufficient features in the dataset.

Question 2 : Can I predict Stock prediction in Qlik ML?
Sample Data:

Date

Symbol

Series

Prev Close

Open

High

Low

Last

VWAP

Volume

Turnover

Trades

Deliverable Volume

%Deliverble

########

MUNDRAPORT

440

770

1050

770

959

962.9

984.72

27294366

2.69E+15

9859619

0.3612

I have derived the below columns in addition to create the ML dataset. But it didn't work yet.

Date	Symbol	Series	Prev Close	Open	High	Low	Last	Close	VWAP	Volume	Turnover	Trades	Deliverable Volume	%Deliverble	Avg CP 5 days	Avg CP Last month	Avg CP LY	Avg CP 5 days/ Avg CP Last Month (P/Q)	Avg CP 5 days/ Avg CP LY (P/R)	Avg CP Last month / Avg CP LY (Q/R)	Avg Vol 5 days	Avg Vol Last month	Avg Vol LY	Avg Vol 5 days/ Avg Vol Last Month (P/Q)	Avg Vol 5 days/ Avg Vol LY (P/R)	Avg Vol Last month / Avg Vol LY (Q/R)	Median	Median - CS	(Median - CS)^2	Sum(SD)/Count	Daily Volatility	Avg SD 5 days	Avg SD Last month	Avg SD LY	Avg SD 5 days/ Avg SD Last Month (AG/AH)	Avg SD 5 days/ Avg SD LY (AG/AI)	Avg SD Last month / Avg SD LY (AH/AI)
########	MUNDRAPORT	EQ	1074.95	1091	1116	1046.3	1078	1066.9	1082.93	845666	9.16E+13		344171	0.41	1085.31	1030.93	1030.93	1.05	1.05	1	1523206	3909740	3909740	0.39	0.39	1	1030.93	-35.97	1293.67	4765.84	69.04	71.82	54.47	54.47	1.32	1.32	1

Thanks in advance.

@KellyHobson

Re: Possible Machine Line learning types we can perform in Qlik Auto ML?

KellyHobson — Thu, 01 Sep 2022 23:57:31 GMT

Hey @Venkadesh_Ponnu ,

Thank you for reaching out in the AutoML forum!

For question #1, AutoML on Qlik Cloud supports binary classification, multi class classification, and regression problems.

The Target you select is used to determine what kind of algorithms to use in the analysis process.

If you select a Target that has only two unique values, algorithms will be used that work best with binary classification problems. For example: customer retention (will my customer leave, yes/no), employee retention (will my employee leave, yes/no), etc.

If you select a Target that is a string value and has more than two unique values, algorithms will be used that work best with multi class classification problems. For example: campaign mix, product recommendation, up-sell opportunity, etc.

If you select a Target that is a number, algorithms will be used that work best with regression problems. For example: how much will this customer purchase, what will be the value of this sale, etc.

Additional information in the User Guide -> https://help.qlik.com/en-US/cloud-services/Subsystems/Hub/Content/Sense_Hub/AutoML/machine-learning-framework.htm

For question #2, it depends on how you are defining you target variable and what question you are asking

Which data attribute/variable are you trying to predict?

For stock data, a more common application is time series forecasting. This capability is built natively in the Qlik Sense line chart.

https://help.qlik.com/en-US/cloud-services/Subsystems/Hub/Content/Sense_Hub/Visualizations/LineChart/timeseries-forecast.htm

ref-> https://towardsdatascience.com/predicting-stock-prices-using-a-keras-lstm-model-4225457f0233

Let me know if you have any additional questions.

Best,

Kelly

Re: Possible Machine Line learning types we can perform in Qlik Auto ML?

Venkadesh_Ponnu — Wed, 07 Sep 2022 12:09:10 GMT

@KellyHobson Thank you very much for your detailed information. It's very nice to reach out to you.