Hi,
Thank you! One quick question? How does the score works in recommendation
engine (
http://predictionio.incubator.apache.org/templates/recommendation/quickstart/)
, there is a rank key in engine.json which value is 10.
Does that mean that highest score is 10?
But My json response shows the score of 11 also, how is it working? Is it
in 100?
Can you please tell me. How the score is working, like what is the highest
score ?
My engine json is like this:
"id": "default",
"description": "Default settings",
"engineFactory": "rec.RecommendationEngine",
"datasource": {
"params" : {
"appName": "MyApp1"
}
},
"algorithms": [
{
"name": "als",
"params": {
"rank": 10,
"numIterations": 5,
"lambda": 0.01,
"seed": 3
}
}
]
}
And the response is something like this
request
<WSGIRequest: GET '/scrapper/'>
response
{'X-Frame-Options': u'SAMEORIGIN',
u'itemScores': [{u'item': u'BM59878', u'score': 11.408240266302734},
{u'item': u'BM62270', u'score': 8.614610809881686},
{u'item': u'BM42896', u'score': 7.696197905320034},
{u'item': u'BM67243', u'score': 7.649859413457619},
{u'item': u'BM65297', u'score': 7.087747727499853},
{u'item': u'BM46953', u'score': 6.920845361111224},
{u'item': u'BM66312', u'score': 6.763293733173957},
{u'item': u'BM67161', u'score': 6.738217253589268},
{u'item': u'BM66971', u'score': 6.630235054477863},
{u'item': u'BM66793', u'score': 6.562655493440761}]}
self
<django.middleware.common.CommonMiddleware object at 0x7f1555eccf50>
On Mon, Mar 27, 2017 at 11:03 AM, Marius Rabenarivo <
[email protected]> wrote:
> You can find more informations about collecting events here :
> http://predictionio.incubator.apache.org/datacollection/eventapi/
>
> 2017-03-27 9:15 GMT+04:00 Marius Rabenarivo <[email protected]>:
>
>> Hi,
>>
>> Yes, the initial events will remain unless you delete them.
>>
>> Regards,
>>
>> Marius
>>
>> 2017-03-27 7:19 GMT+04:00 Vaghawan Ojha <[email protected]>:
>>
>>> Hi,
>>>
>>> I wanted to know how the saved data works inside pio. Let's say I
>>> imported initial events as events.json and then trained the model.
>>>
>>> Later after a month, I again uploaded another file let's say
>>> secondmonth.json. Will that initial events will remain when I run pio train
>>> again?
>>>
>>> How does this works? Could you please brief me a bit.
>>>
>>> Thank you
>>> Vaghawan
>>>
>>
>>
>
