Introduction
This blog covers digital signal processing (DSP), machine learning, time-series analysis, state estimation, and cryptography with Python implementations. This article recommends 12 Japanese-language technical books, organized by level, for readers who want to study each field systematically or keep a solid reference on hand.
Each entry links to the corresponding implementation article on this blog. In this field, alternating between “grasp the overview in an article” and “solidify the framework in a book” is especially effective.
Note: the books below are in Japanese. Titles are given in Japanese with a short English description of what each covers.
1. Digital Signal Processing (DSP)
Signal processing is the core theme of this blog. To learn FFT, filter design, and spectral analysis, these four form the foundation.
- はじめて学ぶディジタル・フィルタと高速フーリエ変換 (CQ Publishing) — A classic introduction that explains the principles of digital filters and the FFT from the ground up. A great companion to The FFT: Theory and Implementation and the Filter Design Guide.
- Python対応 ディジタル信号処理 (Morikita) — DSP fundamentals (sampling, FFT, filter design) with Python code. The best match for this blog’s scipy.signal implementations; pairs well with the DSP × ML Learning Roadmap.
- ディジタル信号処理のエッセンス (Ohmsha) — A compact summary from the sampling theorem to the DFT and Z-transform. Handy for quickly reviewing the math behind Discrete Signal Processing Basics.
- ウェーブレット変換の基礎 (Morikita) — A careful treatment of the wavelet transform for non-stationary signals that FFT/STFT cannot capture. For readers going deeper into the Time-Frequency Analysis Guide.
2. Machine Learning, Bayes, and Deep Learning
- ガウス過程と機械学習 (Kodansha) — A well-regarded book on Gaussian process regression, balancing intuition and mathematics. Read alongside Gaussian Process Regression in Practice.
- ベイズ推論による機械学習入門 (Kodansha) — A systematic introduction to probabilistic models and Bayesian inference. Solidifies the background for Bayesian Optimization and MCMC.
- ゼロから作るDeep Learning (O’Reilly Japan) — The standard book for learning neural networks by implementing them from scratch with only NumPy, no frameworks. The go-to first book for understanding backpropagation and optimization internals; the foundation for SGD / Adam optimization and Self-Attention.
- つくりながら学ぶ! PyTorchによる発展ディープラーニング (Mynavi) — A hands-on PyTorch book covering object detection, segmentation, NLP, and Transformers. A next step after Deep Learning from Scratch for strengthening the implementation skills behind Transformer Time-Series Forecasting and LSTM Time Series.
3. Time-Series Analysis
- Pythonではじめる時系列分析入門 (Kodansha) — Covers the main time-series methods from ARIMA and state-space models to machine learning, with Python code. Corresponds to the ML Time-Series Forecasting Guide and ARIMA.
- 経済・ファイナンスデータの計量時系列分析 (Asakura) — A standard, rigorous text on stationarity, unit roots, ARMA, and state-space models. A step up for deepening the theory behind Autocorrelation and PACF / AR Order Identification.
4. State Estimation and Control
- カルマンフィルタの基礎 (Tokyo Denki University Press) — A standard book that carefully covers the Kalman filter from derivation to implementation. Read together with The Kalman Filter, the Extended Kalman Filter (EKF), and the RTS Smoother.
5. Cryptography and Security
- 暗号技術入門 第3版 秘密の国のアリス (SB Creative) — A classic introduction covering classical ciphers, symmetric and public-key cryptography, hashes, digital signatures, and TLS in plain language with figures. The ideal map to read first for the Cryptography Roadmap.
A Suggested Learning Order
- If you read only one: start with the introductory book in your main field.
- To learn by doing: run an implementation book (e.g. Python対応 ディジタル信号処理, つくりながら学ぶ! PyTorch) alongside this blog’s articles.
- To solidify the theory: use a textbook (e.g. ディジタル信号処理のエッセンス, 経済・ファイナンスデータの計量時系列分析) to back up the math.
Using books for the framework and articles for the implementation — running both in parallel — is the shortest path to mastering this field on your own.
Summary
- We selected 12 books across signal processing, machine learning, time series, state estimation, and cryptography, from introductory to advanced.
- Each book corresponds to an implementation article on this blog; alternating “article for the overview, book for the framework” is effective.
- Start with one introductory book in your main field, then expand to implementation and theory books.
Related Articles
- DSP × ML Learning Roadmap - The map that organizes the articles for each field covered here.
- Digital Filter Design Guide - The article-side entry point for learning filter design.
- ML Time-Series Forecasting Guide - A hub organizing time-series and machine-learning method selection.
- Cryptography Roadmap - The article-side entry point for studying cryptography systematically.
Related Tools
- DevToolBox - Free Developer Tools - Over 90 developer tools including JSON formatter and regex tester.
- CalcBox - Everyday Calculators - Over 60 calculators including statistics and frequency conversion.