def main():
data = dataset_block(load_cancer_dataset(), withOnes=False)
data_train, data_test = split(data)
reg_const = optimize_regularization(data)
theta = svm.train(data_train, c=reg_const)
stats = svm.test(data_test, theta)
print("precision:%6.2f\nrecall:%6.2f\nerror:%6.2f\nf1-score:%6.2f\n" %
(stats.precision(), stats.recall(), stats.error(), stats.f_score()))
print("regularization constant used: %6.2f\n" % reg_const)
def main():
data = dataset_block(load_cancer_dataset(), withOnes=False)
data_train, data_test = split(data)
reg_const = optimize_regularization(data)
theta = svm.train(data_train, c=reg_const)
stats = svm.test(data_test, theta)
print("precision:%6.2f\nrecall:%6.2f\nerror:%6.2f\nf1-score:%6.2f\n" %
(stats.precision(), stats.recall(), stats.error(), stats.f_score()))
print("regularization constant used: %6.2f\n" % reg_const)
def optimize_regularization(data):
reg_best, f1_best = 0, 0
data_train, data_test = split(data)
for d in range(-10, 40):
reg_current = 0.5 ** d
theta = svm.train(data_train, c=reg_current)
stats = svm.test(data_test, theta)
f1_current = stats.f_score()
if f1_best < f1_current:
reg_best, f1_best = reg_current, f1_current
return reg_best
def optimize_regularization(data):
reg_best, f1_best = 0, 0
data_train, data_test = split(data)
for d in range(-10, 40):
reg_current = 0.5**d
theta = svm.train(data_train, c=reg_current)
stats = svm.test(data_test, theta)
f1_current = stats.f_score()
if f1_best < f1_current:
reg_best, f1_best = reg_current, f1_current
return reg_best
from flask import redirect
import download_corpora
import NLProcessor as nlp
#FIX ABOVE 2 LILNES IM NOT SURE WHAT TO DO
from ml import svm
import otherAPIs
import scraper
import searchFunction
import stream
import boto3
import lxml.html
import requests
from requests import get
from goose import Goose
svm.train()
def getSuggestions(query):
url = 'https://api.cognitive.microsoft.com/bing/v5.0/suggestions/?q=' + query
headers = {'Ocp-Apim-Subscription-Key': '854e8088bb8347418e6f934b996487af'}
r = requests.get(url, headers=headers)
results = []
suggestions = r.json()['suggestionGroups']
max = 3
for suggestion in suggestions:
s = suggestion['searchSuggestions']
for term in s:
