I am Praveen Malla, currently working as a Data Scientist with Infosys - a leading IT service provider in India, with 4+ years of experience. I have mentored and recruited a lot of candidates for my company in the data science domain . I have done my B.E in Electronics and communication engineering in BIT, Mesra in 2018. My hobbies are playing badminton and football. For Mentoring Session : https://topmate.io/mallapraveen

My Mentoring Topics

  • Data Science
  • Machine Learning
  • Deep Learning
  • NLP
P.
19.February 2023

Yes session is very helpful He clearly explained everything and provided the good resources for learning I loved the session

D.
24.January 2023

Really helpful session. Got an overall idea about my data career. Very precise and conveyed the whole gist clearly.

Deep Learning with Python, Second Edition
Francois Chollet

Key Insights from "Deep Learning with Python, Second Edition" by Francois Chollet The book elaborates on the fundamental principles and concepts of deep learning. It provides an in-depth understanding about how deep learning models work and how they learn from data. Chollet emphasizes the importance of practical implementation of concepts. The book is full of Python code examples that readers can use for hands-on learning. It covers a range of deep learning architectures including Convolutional Neural Networks (CNNs), Recurrent Neural Networks (RNNs), and Transformers, among others. One of the core concepts discussed by Chollet is how to pre-process data and make it suitable for deep learning models. The book also delves into how to evaluate model performance and how to adjust model parameters to improve their accuracy and efficiency. Chollet discusses how to use Keras, a high-level neural networks API, to simplify the process of designing and training deep learning models. It provides insights into the latest advances in deep learning, including transfer learning, generative adversarial networks (GANs), and self-supervised learning. The book also explores the broader implications of deep learning, including its ethical considerations. Chollet provides guidance on how to tackle real-world problems using deep learning. It includes case studies and application examples from a range of fields including computer vision, natural language processing, and autonomous driving. The book concludes with a look at the future of deep learning, discussing the potential advances and challenges that lie ahead. An In-Depth Analysis of "Deep Learning with Python, Second Edition" Deep Learning with Python, Second Edition is an authoritative guide to deep learning, written by Francois Chollet, the creator of Keras. It is a comprehensive book that provides both theoretical and practical knowledge about deep learning. The book is structured in such a way that it gradually builds up the complexity of topics, making it an excellent resource for both beginners and experienced practitioners. The book commences with an introduction to the fundamental principles and concepts of deep learning. Chollet explains what deep learning is and how it differs from traditional machine learning. He elaborates on how deep learning models are trained using large amounts of data and how they learn to make predictions or decisions without being explicitly programmed to do so. This part of the book is of paramount importance as it lays the foundation upon which the rest of the book is built. Practical implementation of concepts is a recurring theme in the book. Chollet provides numerous Python code examples that readers can experiment with. This hands-on approach facilitates a deeper understanding of the concepts and allows readers to learn by doing. The book covers a wide range of deep learning architectures. Convolutional Neural Networks (CNNs), which are primarily used for image processing tasks, and Recurrent Neural Networks (RNNs), which are used for sequential data such as time series or text, are discussed in detail. In addition, the book also talks about the Transformer architecture, which is the backbone of many state-of-the-art models in natural language processing. Data pre-processing is a crucial step in any machine learning pipeline, and deep learning is no exception. Chollet explains how to prepare data so that it can be effectively used by deep learning models. He covers various techniques such as normalization, tokenization, and vectorization. Evaluating model performance and tuning model parameters are other key aspects discussed in the book. Chollet provides guidance on how to choose the right metrics for evaluation, how to use validation sets, and how to tune hyperparameters to improve model performance. The use of Keras is another significant element of the book. Keras is a high-level neural networks API that is designed to enable fast experimentation with deep learning models. Chollet, being the creator of Keras, provides valuable insights into how to leverage this powerful tool to simplify the process of designing and training deep learning models. The book also provides insights into the latest advances in deep learning. Topics such as transfer learning, generative adversarial networks (GANs), and self-supervised learning are covered. These topics are at the forefront of research in deep learning and have a wide range of applications. Ethical considerations in deep learning are another crucial aspect discussed by Chollet. As deep learning models become increasingly prevalent in our society, it is crucial to consider their ethical implications. Chollet discusses the potential biases in machine learning models and the importance of fairness, transparency, and accountability in AI. The book also provides a practical perspective on deep learning by discussing how to tackle real-world problems. It includes case studies and application examples from a variety of fields including computer vision, natural language processing, and autonomous driving. These examples provide a practical context to the theoretical concepts discussed in the book. Finally, the book concludes with a look at the future of deep learning. Chollet discusses the potential advances in the field, such as the development of more efficient training methods and the creation of more powerful models. However, he also highlights the challenges that lie ahead, including the need for more transparent and explainable models, and the ethical and societal implications of widespread AI adoption. In conclusion, Deep Learning with Python, Second Edition is a comprehensive and accessible guide to deep learning. It combines theoretical knowledge with practical examples, making it an invaluable resource for anyone interested in this field. Whether you are a beginner looking to delve into deep learning or an experienced practitioner seeking to update your knowledge, this book is a must-read.

View
Approaching (Almost) Any Machine Learning Problem
Abhishek Thakur

Key Insights from the Book Practical Approach: The book focuses on practical aspects of machine learning, with an emphasis on real-world applications and problem-solving over theory and math. End-to-End Process: It covers the entire machine learning pipeline, including data gathering, feature engineering, model selection, training, tuning, evaluation, and deployment. Feature Engineering: The book places a lot of importance on feature engineering, which is often overlooked in other machine learning resources. Model Selection: It provides a comprehensive guide on how to choose the right model for your problem and dataset. Model Interpretability: The author emphasizes the importance of model interpretability and how to achieve it, which is crucial for deploying machine learning models in real-world applications. Dealing with Imbalanced Data: The book provides practical strategies for dealing with imbalanced datasets, a common issue in many machine learning problems. Ensembling Techniques: It covers various ensembling techniques, which can greatly improve the performance of your models. Deep Learning: There's a dedicated section on deep learning, including how to design and implement neural networks. Validation Strategies: The book provides insightful guide on how to create robust validation strategies to ensure the generalizability of your models. Competitive Machine Learning: The author shares his experience and tips on competitive machine learning, including Kaggle competitions. In-Depth Analysis and Conclusion "Approaching (Almost) Any Machine Learning Problem" by Abhishek Thakur provides a comprehensive guide to machine learning, focusing on practical problem-solving over theory. The book starts by addressing the end-to-end process of a machine learning project. This is crucial for beginners, as it gives them a concrete roadmap to follow. It also helps experienced practitioners streamline their workflow. The book places a lot of emphasis on feature engineering, an aspect often overlooked in machine learning resources. Feature engineering involves creating new features from the existing data to improve the performance of machine learning models. The author provides numerous practical examples and strategies for feature engineering, which can be applied to a wide range of problems. One of the most valuable sections of the book is on model selection. The author provides a detailed guide on how to choose the right model for your problem, based on the characteristics of your data and the requirements of your application. This is a critical skill, as the choice of model can greatly impact the performance of your machine learning system. The book also covers model interpretability, which is crucial for deploying machine learning models in real-world applications. The author provides various techniques for making your models more interpretable, such as LIME and SHAP. Dealing with imbalanced data is a common issue in machine learning, and the book provides practical strategies for addressing it. These strategies include resampling techniques, cost-sensitive learning, and ensemble methods. The author also covers various ensembling techniques, such as bagging, boosting, and stacking. Ensembling can greatly improve the performance of your models, especially in competitive machine learning. The book includes a section on deep learning, including how to design and implement neural networks. This is a valuable addition, as deep learning is becoming increasingly important in many fields. The book provides insightful guide on how to create robust validation strategies. This is important for ensuring the generalizability of your models, and preventing overfitting. Finally, the author shares his experience and tips on competitive machine learning, including Kaggle competitions. This is a unique aspect of the book, and provides valuable insights for those interested in participating in machine learning competitions. In conclusion, "Approaching (Almost) Any Machine Learning Problem" is a comprehensive and practical guide to machine learning. It covers a wide range of topics, with a focus on real-world applications and problem-solving. Whether you're a beginner or an experienced practitioner, this book will equip you with the knowledge and skills to tackle almost any machine learning problem.

View
Natural Language Processing with Transformers - Building Language Applications with Hugging Face
Lewis Tunstall, Leandro von Werra, Thomas Wolf

Key Facts and Insights Introduction to Natural Language Processing (NLP): The book provides a comprehensive introduction to NLP, explaining how it is used to help computers understand, interpret, and manipulate human language. Deep Dive into Transformers: The authors delve into the details of Transformer models, illustrating their architecture, working mechanism, and their significant role in improving the performance of NLP tasks. Understanding Hugging Face: The book extensively covers Hugging Face, a popular library for NLP tasks, demonstrating how it simplifies the process of implementing Transformer models. Practical Implementations: The authors present various real-world applications of NLP using Hugging Face and Transformers, including text generation, translation, summarization, and sentiment analysis. Training Custom Models: The book explains how to train custom models using Hugging Face, giving readers the skills to apply these techniques to their own unique tasks. Advanced Topics: The book covers advanced topics in NLP, including attention mechanisms, BERT, GPT-2, and fine-tuning models for specific tasks. Exploring Ethical Considerations: The authors discuss the ethical implications of AI and NLP, emphasizing the need for responsible AI development and use. Future of NLP: The final chapters speculate on the future of NLP, keeping in view the rapid advances in the field and potential upcoming developments. In-depth Analysis and Summary Authored by experts in the field - Lewis Tunstall, Leandro von Werra, and Thomas Wolf, this book serves as a comprehensive guide to Natural Language Processing (NLP) using Transformers and the Hugging Face library. The authors have done a commendable job in breaking down complex concepts into digestible content, making the book accessible to both beginners and experienced practitioners in the field. The book begins with a robust introduction to NLP, outlining its importance in enabling computers to understand and interact with human language. This foundation provides readers with the necessary context to appreciate the intricate workings of NLP. The heart of the book dives into the fascinating world of Transformer models. These models have revolutionized NLP tasks with their ability to handle long-range dependencies in text, solving problems inherent in previous architectures like RNNs and LSTMs. The authors meticulously explain the architecture of these models, focusing on their key components - self-attention mechanisms and positional encodings. They further delve into the concept of attention mechanisms, examining how they allow models to focus on relevant information while processing language data. Hugging Face, a library that has emerged as a game-changer in the NLP landscape, is covered extensively in the book. The authors illustrate how Hugging Face simplifies the implementation of Transformer models and democratizes AI by providing pre-trained models that can be fine-tuned for specific tasks. The book provides practical examples of various NLP applications using Hugging Face and Transformers, including text generation, translation, summarization, and sentiment analysis. This practical approach helps readers understand and apply the concepts they learn. Training custom models is another major topic covered in the book. Readers are guided through the process of training their own models using Hugging Face, equipping them with the skills to tailor models to their unique tasks. This is further complemented by a discussion on more advanced topics in NLP, such as BERT and GPT-2, which have made significant contributions to the field. The authors also take time to explore the ethical implications of AI and NLP, a crucial yet often overlooked aspect in the rush of technological advancements. They emphasize the importance of responsible AI development and usage, discussing potential biases in NLP models and exploring ways to mitigate them. In conclusion, the book speculates on the future of NLP, considering the rapid advances in the field and the potential for further breakthroughs. This forward-looking perspective provides readers with a broader context for the ongoing developments in NLP. In essence, this book provides a comprehensive exploration of NLP using Transformers and the Hugging Face library, making it an invaluable resource for anyone interested in the field. It balances theoretical understanding with practical implementation, providing readers with a well-rounded understanding of the subject. The authors' expertise and clear explanation make complex concepts accessible, offering readers the knowledge and skills to apply these techniques in their own work. From a broader perspective, the book also prompts readers to consider the ethical implications of their work, fostering a more responsible approach to AI development.

View
Designing Data-Intensive Applications - The Big Ideas Behind Reliable, Scalable, and Maintainable Systems
Martin Kleppmann

Key Facts and Insights The book explores the underlying principles of data systems and how they are used to build reliable, scalable, and maintainable applications. It outlines the importance of distributed systems in handling data-intensive applications and how to deal with the challenges associated with them. The book emphasizes on the trade-offs involved in choosing particular data structures, algorithms, and architectures for data-intensive applications. It provides a detailed explanation of the three main components of data systems: storage, retrieval, and processing. It presents an in-depth understanding of consistency and consensus in the context of distributed systems. The book discusses various data models, including relational, document, graph, and many more, along with their suitable use cases. It also examines the concept of stream processing and batch processing, their differences, and when to use each. It underlines the significance of maintaining data integrity and the techniques to ensure it. It offers comprehensive coverage of the replication and partitioning strategies in distributed systems. The book provides a balanced view of various system design approaches, explaining their strengths and weaknesses. Lastly, the book does not recommend one-size-fits-all solutions. Instead, it equips the reader with principles and tools to make informed decisions depending on the requirements of their projects. In-Depth Analysis of the Book "Designing Data-Intensive Applications" by Martin Kleppmann is a comprehensive guide to understanding the fundamental principles of data systems and their effective application in designing reliable, scalable, and maintainable systems. It provides an exhaustive account of the paradigms and strategies used in data management and their practical implications. Understanding Data Systems The book begins by introducing the basics of data systems, explaining their role in managing and processing large volumes of data. It delves into the three main components of data systems: storage, retrieval, and processing. Each component is explored in detail, providing the reader with a clear understanding of its functionality and importance in a data system. Data Models and Query Languages The book delves into the various data models used in data-intensive applications, such as relational, document, and graph models. It provides a comparative analysis of these models, highlighting their strengths and weaknesses, and the specific use cases they are best suited for. Additionally, it discusses the role of query languages in data interaction, explaining how they facilitate communication between the user and the data system. Storage and Retrieval The book explains the techniques and data structures used for efficiently storing and retrieving data. It underlines the trade-offs involved in choosing a particular approach, emphasizing the importance of taking into account the specific requirements of the application. Distributed Data The book delves into the complexities of distributed data. It outlines the significance of distributed systems in handling data-intensive applications and discusses the challenges associated with them, such as data replication, consistency, and consensus. It also provides solutions to these challenges, equipping the reader with strategies to effectively manage distributed data. Data Integrity The book underscores the significance of maintaining data integrity. It provides an in-depth understanding of the concept and discusses techniques to ensure it, such as atomicity, consistency, isolation, and durability (ACID) and base properties. Stream Processing and Batch Processing The book examines the concept of stream processing and batch processing. It discusses their differences, the challenges associated with each, and the scenarios where one would be preferred over the other. Conclusion In conclusion, "Designing Data-Intensive Applications" is a comprehensive guide that provides readers with a deep understanding of data systems. It equips them with the knowledge to make informed decisions when designing data-intensive applications, based on the specific requirements of their projects. The book's strength lies in its balanced view of various system design approaches, offering a holistic understanding of the dynamics involved in managing data. It is an essential read for anyone seeking to delve into the world of data systems.

View