MathML assignments + kaggle challenges

.gitmodules

@@ -0,0 +1,3 @@
+[submodule "assMath/probStat/aakashks/kalman_assignment"]
+	path = assMath/probStat/aakashks/kalman_assignment
+	url = https://github.com/aakashks/kalman_filter_assignment

README.md

@@ -18,5 +18,11 @@ https://www.kaggle.com/datasets/ahmedmohameddawoud/ecommerce-ab-testing
 
 Look at the code contributions in the link to get an idea of how its implemented.
 
-## MATH-2
+## MATH & ML
+
+The assignment has been divided into 3 parts and you are expected to:
+### -Complete atleast one part by 21st - https://hackmd.io/1GLPNImwQbCbYNeIHc5Bww?view
+### -Complete atleast two parts by 25th -https://drive.google.com/file/d/1yD35uuig5mYxQlbi3mnPl0htKtcCheHp/view?usp=sharing
+### -Complete the entire assignment by 28th - https://hackmd.io/rRMAoU02RbyASZGMrubPPA?view
+
 ## ML-0

aakashks/Git-Bash_aakashks/ML_automation/ML_script.py

@@ -0,0 +1,38 @@
+import argparse
+import pandas as pd
+from sklearn.ensemble import RandomForestClassifier
+
+parser = argparse.ArgumentParser()
+parser.add_argument('-d', '--train-data')
+parser.add_argument('-t', '--test-data')
+parser.add_argument('-f', '--target-feature')
+parser.add_argument('-p', '--parameters', nargs=2, type=int)
+
+args = parser.parse_args()
+
+# Assuming that the data file train_data is preprocessed
+train_data = pd.read_csv(args.train_data_path)
+test_data = pd.read_csv(args.test_data_path)
+
+# Separating X and y
+y = train_data[args.target_feature]
+X = train_data.drop(args.target_feature)
+X_test = test_data.drop(args.target_feature)
+
+model = RandomForestClassifier(
+    n_estimators=args.parameters[0],
+    max_depth=args.parameters[1],
+    random_state=42
+)
+
+print('Fitting model')
+model.fit(X, y)
+
+print('generating predictions')
+predictions = model.predict(X_test)
+output_predictions = pd.DataFrame(
+    {'ID': test_data.index, args.target_feature: predictions}
+)
+
+print('Saving predictions in a file')
+output_predictions.to_csv('./submissions.csv', index=False)

aakashks/Git-Bash_aakashks/ML_automation/README.md

@@ -0,0 +1,7 @@
+# ML automation script using bash and argparse
+
+The script is supposed to take preprocessed csv file paths and fit the model with user defined parameters
+then the predictions are output to submissions file and the logs are logged into output.txt.
+
+It uses a Random Forest Classifier on the data. User needs to specify the n_estimators and max_depth
+parameters of the sklearn model.

aakashks/Git-Bash_aakashks/ML_automation/automate_script.sh

@@ -0,0 +1,16 @@
+#!/bin/bash
+
+output_path=./output.txt
+
+# Getting input from user
+echo 'path of training data: '
+read train_data_path
+echo 'path of test data: '
+read test_data_path
+
+read -p 'target feature name: ' target_feature
+read -p 'model parameters: ' n_estimators max_depth
+
+# Running python script
+python ML_script.py -d $train_data_path -t $test_data_path \
+ -f $target_feature -p $n_estimators $max_depth &> $output_path

aakashks/Git-Bash_aakashks/README.md

@@ -0,0 +1,8 @@
+## task 1 & 2
+in respective .md files
+
+## task 4
+[automate_script](automate_script.sh)
+
+## task 5
+done

aakashks/Git-Bash_aakashks/bandit.md

@@ -0,0 +1,68 @@
+# bandit solutions
+
+```bash
+ssh [email protected] -p 2220
+```
+
+0. cat readme
+1. cat ./-
+2. cat "spaces in this filename"
+3.
+
+```bash
+ls -al
+cat .hidden
+```
+
+4.
+
+```bash
+ll
+find . -type f | xargs file | grep text
+cat ./-file07
+```
+
+5. find . -type f -size 1033c ! -executable
+6. find / -type f -size 33c ! -executable -user bandit7 -group bandit6 2>/dev/null
+
+for this one handling error stream is very important as it was displaying errors with the required output. hence 2> :error stream, /dev/null is what prints null
+
+7. cat data.txt | grep millionth
+8. sort data.txt | uniq -u
+9.  strings data.txt | grep ===
+10. base64 --decode data.txt | cat
+11. cat data.txt | tr 'a-zA-Z' 'n-za-mN-ZA-M'
+12.
+
+```bash
+# make folder
+mkdir /tmp/temp
+mkdir /tmp/t11
+cd /tmp/t11
+cp ~/data.txt .
+mv data.txt hex_dump.txt
+
+# convert hex dump to ASCII
+xxd -r hex_dump.txt > compressed.bin
+
+# now we'll check zip type with
+file filename
+
+# and use zcat, bzcat to get tar file
+# for tar file we use tar -xvf 
+# after many times we get the text
+
+zcat compressed.bin > compressed2
+bzcat compressed2 > compressed3
+tar -xvf compressed4
+```
+
+13. ssh -i sshkey.private [email protected] -p 2220
+
+14. 
+cat /etc/bandit_pass/bandit14
+echo fGrHPx402xGC7U7rXKDaxiWFTOiF0ENq | nc localhost 30000
+
+15.echo jN2kgmIXJ6fShzhT2avhotn4Zcka6tnt | openssl s_client -connect localhost:30001 -ign_eof
+
+password for 16 => JQttfApK4SeyHwDlI9SXGR50qclOAil1