CIS AIML Beginners Series Part 1
2 likes696 views
This document outlines a beginner's series on artificial intelligence, machine learning, and deep learning, emphasizing foundational concepts, practical applications, and the importance of understanding the relevant technologies. It introduces essential tools and libraries like TensorFlow, Jupyter Notebook, and various algorithms useful for machine learning. The session concludes with a preview of upcoming topics related to coding and advanced ML techniques.
CIS AIML Beginners Series Part 1
- 2. Getting closer to BEGINNER’S SERIES – PART 1 BY MILAN 2 C ouncilofInform ation Security
- 3. What’s the Agenda? Clearing our Basics Understanding AI ML DL There is more to it Setting up your Weapon Understand your Weapon Know the Libraries Run the Code What you want next? 3 C ouncilofInform ation Security
- 4. Know your Trainer 4 Milan Singh Thakur Artificial Intelligence & Machine Learning Evangelist Executive Director – Future Technologies with CIS OWASP Mobile Project Global Leader International speaker, trainer, learner C ouncilofInform ation Security
- 5. What is the Confusion? 5 Where to start from? Which course is better? Should I take online/offline training? Which direction will it take my career? Do I need to work with lot of math? Do I need to be a programmer? Will my colleagues understand what I am doing? C ouncilofInform ation Security
- 7. AI, ML and DL..?? Artificial Intelligence Machine Learning Deep Learning 7 C ouncilofInform ation Security
- 8. Why to use ML now? Too much of data available Computational power is increasing Algorithmic research is breaking innovation barriers Increasing support and interest of industries ML + Cyber Security is, “The Endless Game” IoT and IIoT needs to be smarter Hackers have weaponized ML, when will you? 8 C ouncilofInform ation Security
- 9. Applications of ML 9 C ouncilofInform ation Security
- 11. Ex: Spam Filter using ML 11 C ouncilofInform ation Security
- 12. Ex: Static Malware Analysis using ML 12 C ouncilofInform ation Security
- 13. Check if you are “Still Awake” 13 I want to calculate my share value. Which category? I want to understand what my customers are expecting in next product? My defense system drone has to identify the enemy and execute? My bank detects ATM fraud very quickly, how? Face unlock on my mobile? C ouncilofInform ation Security
- 14. Leveraging ML in Industry Predictive maintenance Demand forecasting Workforce management Consumer buying patterns and recommendations Device monitoring, incident reporting, advanced analytics Airbus has launched “Factory of Future” – a Boeing creates nearly half terabyte data during every flight Bentley uses it for quality vs customer satisfaction prediction IBM Watson is extensively used in Healthcare industry for patient data analysis, diagnostics and help doctors for decision making 14 C ouncilofInform ation Security
- 15. How you can get into Machine Learning? Get into basic courses by Andrew NG Install jupyter notebook – it has all you need (Yolo, PIL, pandas, sk-learn & more) Run sample code: iris flower, object detection Understand you data Understand how labelling of data is done Look into how noise is removed Be wise to choose right algorithm Validate the output 15 C ouncilofInform ation Security
- 16. Setting up you environment http://jupyter.org/install Download and install Anaconda with python3 Download and install jupyter notebook Start the notebook server $Jupyter notebook Visit http://localhost:8888 for Notebook dashboard 16 C ouncilofInform ation Security
- 17. Understanding Jupyter Notebook Dash Demo 17 C ouncilofInform ation Security
- 18. What Libraries to focus on? TensorFlow: Originally developed by researchers and engineers from the Google Brain team within Google’s AI organization, it comes with strong support for machine learning and deep learning and the flexible numerical computation core is used across many other scientific domains. Numpy: Primarily used for scientific computing, has a powerful N-dimensional array object, useful linear algebra, Fourier transform, and random number capabilities OpenCV: Multi-core processing, hardware acceleration, real-time applications Usage ranges from interactive art, to mines inspection, stitching maps on the web or through advanced robotics 18 C ouncilofInform ation Security
- 19. Pandas: Library providing high-performance, easy-to-use data structures and data analysis tools for the Python programming language PIL – Pilow The Python Imaging Library adds image processing capabilities to your Python interpreter Yolo: You Only Look Once (YOLO), real-time object detection system based on a Pascal Titan X it processes images at 30 FPS and has a mAP of 57.9% on COCO test-dev Scikit-learn: Simple and efficient tools for data mining and data analysis, Accessible to everybody and reusable in various contexts. Built on NumPy, SciPy, and matplotlib StatsModels matplotlib 19 C ouncilofInform ation Security
- 20. What are my Algorithms? 20 Classification SVM Nearest Neighbors Random Forest Regression SVR Ridge Regression Clustering K-Means Spectral Clustering Mean shift Dimensionality Reduction PCA Feature selection Non-negative matrix factorization Model Selection Grid search Cross validation metrics Preprocessing Preprocessing Feature extraction C ouncilofInform ation Security
- 21. Run Run Run…!! Let’s run our first code of ML See Object detection video using Yolo 21 C ouncilofInform ation Security
- 22. What’s in Next Session (Part 2)…!! Jumping into Libraries Writing code for object detection Natural Language Processing (NLP) Neural Networks SVM Deep dive into Algorithms Decision Trees and Forests 22 C ouncilofInform ation Security
- 23. “ ” A Vision to Secure the Nation COUNCIL OF INFORMATION SECURITY Thank you for being part of this session. Your participation and feedback is valuable for us. Email us at: info@aimlglobal.org 23 C ouncilofInform ation Security