def initialize(self):
if not self.train_data:
self.train_data = process(self.source)
self.train_data.extract_info()
self.maxstring = len(self.train_data.genotype[0]) - 1
self.total = self.train_data.intrusion + self.train_data.normal
print("---------------")
print("Initializing Random Population")
for i in range(0, self.popsize):
self.population.append(DNA(self.maxstring))
print("Operation Complete")
print("---------------")
import pandas as pd
from sklearn.feature_extraction.text import CountVectorizer, TfidfTransformer
from sklearn.pipeline import Pipeline
from Preprocessing import process
from xgboost import XGBClassifier
import numpy
from Preprocessing import parse_impression
### PROCESSING (X = reports, Y = y)
CTIMEreports = pd.read_csv("data/CTIME/CTIMEFinalLabel.csv")
AllReports = pd.read_csv("data/CTIME/CTIMEDataset.csv")
processed = []
for report in CTIMEreports['CTReport']:
report = parse_impression(report)
rep = ' '.join(process(report))
processed.append(rep)
CTIMEreports['PROCESSED'] = (processed)
processed = []
for report in AllReports['CTReport']:
report = parse_impression(report)
rep = ' '.join(process(report))
processed.append(rep)
AllReports['PROCESSED'] = processed
### BUILD MODEL (can change fit & predict functions to train on 'CTReport' (unprocessed), 'PROCESSED', 'FINDINGS' (processed), IMPRESSION' (processed)
text_clf = Pipeline([
('vect',
CountVectorizer(stop_words='english',
ngram_range=(1, 2),
from Preprocessing import process
test = process(source="KDDTest+.txt")
test.extract_info()
def predict(individual, specimen):
w1 = 0.239
w2 = 0.181
w3 = 0.215
w4 = 0.162
w5 = 0.069
w6 = 0.064
w7 = 0.055
w8 = 0.015
outcome = 0
if individual[0:7] == specimen[0:7]:
outcome += w1
if individual[7:17] == specimen[7:17]:
outcome += w2
if individual[17:29] == specimen[17:29]:
outcome += w3
if individual[29:31] == specimen[29:31]:
outcome += w4
if individual[31:35] == specimen[31:35]:
outcome += w5
if individual[35:45] == specimen[35:45]:
outcome += w6
if individual[45:52] == specimen[45:52]:
outcome += w7