how can we feed multiple stock data and get multiple output? - machine-learning

Question

Usually, we do predictions on data for example USD/GBP which consists of columns HIGH, LOW, OPEN, CLOSE. This is only a prediction on a single dataset. And my question is how can I do prediction on multiple data such as USD/GBP, EUR/USD, XAU/USD, USD/JPY. Feeding all this data to a single model and getting predictions on all this dataset i.e multiple outputs. Is this possible?? Thank you.

Answer 1

If I have got your question correctly what you want is to feed in data to a model which has two models in it and run on one model (model1 or model2) seperately according to the data fed into it. For this you can create a model class with the minimal interface which will create a model which will select based on the value of the feature
As an illustration:
import pickle as pkl
from sklearn.linear_model import LinearRegression
from sklearn.preprocessing import LabelEncoder
#Combined of both df1 and df2
data= {
'HIGH': [1.0, 1.0, 0.0, 1.0, 1.0, 0.0, 0.0, 1.0, 1.0, 1.0, 0.0, 0.0],
'LOW': [0.0, 0.0, 0.0, 1.0, 1.0, 1.0, 2.0, 2.0, 2.0, 3.0, 3.0, 3.0],
'OPEN': [1.0, 2.0, 3.0, 1.0, 2.0, 3.0, 1.0, 2.0, 3.0, 1.0, 2.0, 3.0],
'END' : [3.0, 1.0, 3.0, 4.0, 2.0, 3.0, 1.0, 2.0, 6.0, 1.0, 2.0, 1.0]
'target_price': [0.0, 2.0, 1.5, 0.0, 5.1, 4.0, 0.0, 1.0, 2.0, 0.0, 2.1, 1.5]
}
df= pd.DataFrame(data)
X, y=df.iloc[:, :-1], df.iloc[:, -1]
X= X.astype('float32')
# create two models
model1= LinearRegression()
model2 = LinearRegression()
ser_model1= X['x']==0.0
model1.fit(X[ser_model1], y[ser_model1])
model2.fit(X[~ser_model1], y[~ser_model1])
# define a class that mocks the model interface
class CombinedModel:
def __init__(self, model1, model2):
self.model1= model1
self.model2= model2
def predict(self, X, **kwargs):
ser_model1= X['x']==0.0
return pd.concat([
pd.Series(self.model1.predict(X[ser_model1]), index=X.index[ser_model1]),
pd.Series(self.model2.predict(X[~ser_model1]), index=X.index[~ser_model1])
]
).sort_index()
# create a model with the two trained sum models
# and pickle it
model= CombinedModel(model1, model2)
model.predict(X)
with open('model.pkl', 'wb') as fp:
pkl.dump(model, fp)
model= model1= model2= None
# test load it
with open('model.pkl', 'rb') as fp:
model= pkl.load(fp)
model.predict(X)
OR if you want to ensemble
then use VotingClassifier
The idea behind the voting classifier implementation is to combine conceptually different machine learning classifiers and use a majority vote or the average predicted probabilities (soft vote) to predict the class labels