I use predictionIO, build and trained a model for complimentary purchase template engine: https://github.com/PredictionIO/template-scala-parallel-complementarypurchase
Now I need to evaluate the results. But the evaluation part in predictionio docs is missing. Could someone please give me an advice, which metrics should I use and which evaluation techniques?
Thanks a lot.
Related
I'm a beginner.
I'm trying to research how the behavior of a vehicle changes if you set an estimated time of arrival for each vehicle using sumo and veins, and drive according to that time.
Of course, I think it would be difficult to implement this without traffic conditions and a smart navigation system, but I'm thinking of giving it a try.
Now, when I implement it, I want to add an attribute to that is not provided in sumo.
I checked the sumo file and it seems to be set in the C++ file, but I couldn't find where I should add it.
Also, I think I need to make some changes to the veins to process the data after adding it in sumo, where can I look to learn more?
Here are the tools I'm using
sumo (1.8.0) veins (5.1) omnet++ (5.6.2)
I am sorry, but I would appreciate it if someone could push this for me.
If you want to add arbitrary attributes to vehicles, this can be done using what SUMO calls "Generic Parameters". See https://github.com/eclipse/sumo/blob/v1_8_0/docs/web/docs/Simulation/GenericParameters.md for a quick explanation, see https://github.com/sommer/veins/blob/veins-5.1/src/veins/modules/mobility/traci/TraCICommandInterface.h#L167 for a way to access them from OMNeT++.
I am working in a project with Informatica Data Integration Hub. We are evaluating the idea to use FILTERS in the subscription site:
Basic Filter expressions
Advanced Filter Expressions
In the following picture I'm using some examples of basic expressions ( just for testing purposes, don't think about the meaning of those basic filters, they have no meaning at all ).
I cannot find anything in the documentation regarding which syntax should I follow for these advanced filter expressions. What I read in the documentation is that the filter conditions in advanced expressions should follow the rules of "Powercenter Expressions"
Can anyone tell me where I can find the documentation about the syntax of those expressions ?
Does anyone know why I can put whatever I want in the advanced expression field and it is not even evaluating for syntax purposes ?
Thank you
Here's a the filtering condition doc briefly covering the topic.
Here's a document on working with expressions in Informatica.
Basically it's all pretty Excel-like. You pick a function, put arguments in brackets and that's it.
Now, the difference between basic and advanced filtering is, that in basic mode you put several conditions that need to be met at once, as Informatica is using AND to combine them. In case you need more anvanced AND/OR logic, you need to use the advanced expressions and write the complete syntax on your own.
Its difficult to give more details as I haven't even found a clue in OBEO Designer Community to model OCL constraints.
Anyone has a idea how to do this?
Thanks
You should be able to add Validation rules with a constraint expressed in AQL -- Acceleo Query Language (a flavor of OCL) directly in the .odesign file.
See thecorresponding part of the tutorial.
I would like to know if the WSO2 ML implement Cross-Validation and Grid Search for best model selection.
Presently, (as of version 1.1.0) WSO2 Machine Learner does not have a direct method for hyper-parameters optimization. As mentioned in your question, we are planning to include Random Search and Grid Search in one of the upcoming releases. In order to track the progress of this process, I have created a public JIRA [1]. So when the new feature is ready I will notify you via this SO Question.
Next, let me briefly describe cross-validation process we use in WSO2 Machine Learning server. In the third step of the ML Wizard of the ML Server, you can set the training data fraction (please see the attached screen shot).
So let's say you pick 0.7 of your data for training. Then, model building process will use 70% of your data for training and rest of the dataset (i.e. 30%) will be used for cross-validation. As you might recognize this a most basic approach for cross-validation and it is not particularly suitable for small datasets. So in upcoming releases, we are planning to include K-fold cross-validations [2] in addition to the currently available cross-validation method.
Yandi, if you need further help regarding this question or anything related to our product please let me know.
Thanks,
Upul
[1] https://wso2.org/jira/browse/ML-313
[2] https://en.wikipedia.org/wiki/Cross-validation_(statistics)#k-fold_cross-validation
I recently came to a term in one of my email communicatons with my supervisor.Since I am beinging doing a data-mining project on facebook user profile,and he said I should being collecting groud-truth data.
I am very new to this term and I searched online for it,but found very few results about it in data-mining sense.
Could anyone give me an example of what this groud-truth data is in a data-mining task pleae?
Thank you very much.
Ground-truth is data annotated (generally by human) known to be sure at 100%.
It's used to train algorithm since it's what you expect the algorithm to give you.