- change the format of output from text to dictionary
- unify parameters of binary classification and
multiclass classification
Change-Id: I384ccdb06f3cbda44defe4ab40cd66da6ab09231
Enable NaiveBayes to support a text format using tf-idf.
This feature allows user to create following prediction models.
- model detects whether it is a spam mail or not
- model predicts whether it is a review of goodwill or not
- model detects what language a document is written in
implements blueprint : support-text-format
Change-Id: Ib016c587ab821faa7518f6face76d57883dab5ab
Meteos only supports DecisionTreeRegression
in tree models until now.
This patch add support models as below.
- DecisionTreeClassification
- RandomForestRegression
- RandomForestClassification
implements blueprint add-support-models
Change-Id: Ia41b852c1cc6ffa10033b3a6140a969cf3c6f716
Add a common class named Regression Model and
add Ridge Regression Model.
And add some parameters which are used commonly
in Regression Model.
implements blueprint add-support-models
Change-Id: I4648bf2009ca4910014d3d0afd7496359f48131e
In machine learning, user splits the dataset in general.
One is for creating a prediction model,
the other is for evaluating model.
Add a split dataset command to make it possible.
implements blueprint split-dataset
Change-Id: Idb323a1240d790d5628f9eb31cba8d3a29ad64e8
- accept string format in model parameters for meteos-ui
- modify value list which are returned by meteos-api
Change-Id: I10b270d3f33aaf470466f6daa96f1756739d723e
In Machine Learning,
Prediction model needs to be evaluated accuracy.
Add a "meteos model-evaluation" command to allow users
to evaluate their models which they created.
implements blueprint evaluate-model
Change-Id: I35b9322d423a224311541f940ecf985870d7c396
This patch allows user to load a Prediction Model
in advance for online prediction.
In online prediction,
user can retrieve a predicted data immediately.
implements blueprint online-prediction
Change-Id: I03c3a1fe3cd5bbc09116729647c2efd508259987
A spark.mllib provides a parallel implementation
of FP-growth which is a popular algorithm
to mining frequent itemsets.
So, Meteos should support this model.
This model allows user to find
frequent word sets from text data.
implements blueprint support-fp-growth
Change-Id: If44578acf586ed7a68afc1bd1a7cceaa1c036bd7
Hiroyuki Eguchi