piEsposito

import torch
import torch.nn as nn
import torch.nn.functional as F
import torch.optim as optim
import numpy as np

from blitz.modules import BayesianLinear
from blitz.utils import variational_estimator

from sklearn.datasets import load_boston

piEsposito

X, y = load_boston(return_X_y=True)
X = StandardScaler().fit_transform(X)
y = StandardScaler().fit_transform(np.expand_dims(y, -1))

X_train, X_test, y_train, y_test = train_test_split(X,
                                                    y,
                                                    test_size=.25,
                                                    random_state=42)

piEsposito

X, y = load_boston(return_X_y=True)
X = StandardScaler().fit_transform(X)
y = StandardScaler().fit_transform(np.expand_dims(y, -1))

X_train, X_test, y_train, y_test = train_test_split(X,
                                                    y,
                                                    test_size=.25,
                                                    random_state=42)

piEsposito

X, y = load_boston(return_X_y=True)
X = StandardScaler().fit_transform(X)
y = StandardScaler().fit_transform(np.expand_dims(y, -1))

X_train, X_test, y_train, y_test = train_test_split(X,
                                                    y,
                                                    test_size=.25,
                                                    random_state=42)

piEsposito

X, y = load_boston(return_X_y=True)
X = StandardScaler().fit_transform(X)
y = StandardScaler().fit_transform(np.expand_dims(y, -1))

X_train, X_test, y_train, y_test = train_test_split(X,
                                                    y,
                                                    test_size=.25,
                                                    random_state=42)

piEsposito

@variational_estimator
class BayesianRegressor(nn.Module):
    def __init__(self, input_dim, output_dim):
        super().__init__()
        #self.linear = nn.Linear(input_dim, output_dim)
        self.blinear1 = BayesianLinear(input_dim, 512)
        self.blinear2 = BayesianLinear(512, output_dim)
        
    def forward(self, x):
        x_ = self.blinear1(x)

piEsposito

def evaluate_regression(regressor,
                        X,
                        y,
                        samples = 100,
                        std_multiplier = 2):
    preds = [regressor(X) for i in range(samples)]
    preds = torch.stack(preds)
    means = preds.mean(axis=0)
    stds = preds.std(axis=0)
    ci_upper = means + (std_multiplier * stds)

piEsposito

regressor = BayesianRegressor(13, 1)
optimizer = optim.SGD(regressor.parameters(), lr=0.001)
criterion = torch.nn.MSELoss()

ds_train = torch.utils.data.TensorDataset(X_train, y_train)
dataloader_train = torch.utils.data.DataLoader(ds_train, batch_size=16, shuffle=True)

ds_test = torch.utils.data.TensorDataset(X_test, y_test)
dataloader_test = torch.utils.data.DataLoader(ds_test, batch_size=16, shuffle=True)

piEsposito

import pandas as pd
import numpy as np

import torch
import torch.nn as nn
import torch.optim as optim
import torch.nn.functional as F

from blitz.modules import BayesianLSTM
from blitz.utils import variational_estimator

piEsposito

#importing the dataset
amazon="data/AMZN_2006-01-01_to_2018-01-01.csv"
ibm="data/IBM_2006-01-01_to_2018-01-01.csv"
df = pd.read_csv(ibm)

#scaling and selecting data
close_prices = df["Close"]
scaler = StandardScaler()

close_prices_arr = np.array(close_prices).reshape(-1, 1)