دانلود کتاب روش‌های کمی آماری در امور مالی

I began writing this book with the encouragement and active support of my colleagues, including young professionals joining financial firms who frequently seek my recommendations on how to review all relevant statistical theory that is useful for financial modeling used at the workplace. There are reputed textbooks on individual topics such as regression, time series, and generalized linear models, but few covering the complete spectrum of statistical methods used in quantitative modeling in finance. This book was written to address that need and to provide a resource on practical quantitative modeling tasks that can be tackled using statistical methods. While most examples are drawn from the field of finance, students and practitioners from other disciplines may find it useful for understanding the concepts and identifying analogies to the modeling tasks they work on. This work strikes a balance between theory and practice, providing an overview of theory and illustrating it with practical examples. This book assumes the reader is familiar with Python programming. Knowledge of libraries such as statsmodels and sklearn is not required. During the course of reading this book, the reader will acquire a synoptic understanding of frequently used APIs available for the model implementations supported by these libraries. I welcome feedback from readers on what they found helpful in the book and where there is scope for improvement. A text focused on evolving practical concepts must necessarily evolve in future editions to cover new applications and models. This is particularly relevant with the increasing synergy between machine learning and data science. Statistical concepts such as entropy, cross-entropy, maximum likelihood, Gini impurity index, and Bayesianmethods have become the vanguard of a new generation of artificial intelligence models, such as Glove (global vectors for word representation in NLP), double-Q learning in reinforcement learning (crossentropy), and random forests (Gini impurity index), to name a few. On a parallel front, statistical models have been deployed as benchmarking tools during the calibration and model fitting phase of machine learning algorithms and as tools for ensuring ongoing quality controls during postproduction deployment.