The PAST decade has seen an explosion of machine learning research and appli-
cations; especially, deep learning methods have enabled key advances in many
applicationdomains,suchas computervision,speechprocessing,andgameplaying.
However, the performance of many machine learning methods is very sensitive
to a plethora of design decisions, which constitutes a considerable barrier for
new users. This is particularly true in the booming field of deep learning, where
human engineers need to select the right neural architectures, training procedures,
regularization methods, and hyperparameters of all of these components in order to
make their networks do what they are supposed to do with sufficient performance.
This process has to be repeated for every application. Even experts are often left
with tedious episodes of trial and error until they identify a good set of choices for
a particular dataset.
標簽：
Auto-Machine-Learning-Methods-Sys
tems-Challenges
上傳時間：
2020-06-10
上傳用戶：shancjb
Pattern recognition has its origins in engineering, whereas machine learning grew
out of computer science. However, these activities can be viewed as two facets of
the same field, and together they have undergone substantial development over the
PAST ten years. In particular, Bayesian methods have grown from a specialist niche to
become mainstream, while graphical models have emerged as a general framework
for describing and applying probabilistic models. Also, the practical applicability of
Bayesian methods has been greatly enhanced through the development of a range of
approximate inference algorithms such as variational Bayes and expectation propa-
gation. Similarly, new models based on kernels have had significant impact on both
algorithms and applications.
標簽：
Bishop-Pattern-Recognition-and-Ma
chine-Learning
上傳時間：
2020-06-10
上傳用戶：shancjb
Artificial Intelligence (AI) has undoubtedly been one of the most important buz-
zwords over the PAST years. The goal in AI is to design algorithms that transform com-
puters into “intelligent” agents. By intelligence here we do not necessarily mean an
extraordinary level of smartness shown by superhuman; it rather often involves very
basic problems that humans solve very frequently in their day-to-day life. This can
be as simple as recognizing faces in an image, driving a car, playing a board game, or
reading (and understanding) an article in a newspaper. The intelligent behaviour ex-
hibited by humans when “reading” is one of the main goals for a subfield of AI called
Natural Language Processing (NLP). Natural language 1 is one of the most complex
tools used by humans for a wide range of reasons, for instance to communicate with
others, to express thoughts, feelings and ideas, to ask questions, or to give instruc-
tions. Therefore, it is crucial for computers to possess the ability to use the same tool
in order to effectively interact with humans.
標簽：
Embeddings
Processing
Language
Natural
in
上傳時間：
2020-06-10
上傳用戶：shancjb
PrefaceDuring the PAST years, there has been a quickly rising interest in radio access technologies for providingmobile as well as nomadic and fixed services for voice, video, and data. The difference indesign, implementation, and use between telecom and datacom technologies is also becoming moreblurred. One example is cellular technologies from the telecom world being used for broadband dataand wireless LAN from the datacom world being used for voice-over IP.Today, the most widespread radio access technology for mobile communication is digital cellular,with the number of users passing 5 billion by 2010, which is more than half of the world’s population.It has emerged from early deployments of an expensive voice service for a few car-borne users,to today’s widespread use of mobile-communication devices that provide a range of mobile servicesand often include camera, MP3 player, and PDA functions. With this widespread use and increasinginterest in mobile communication, a continuing evolution ahead is foreseen.This book describes LTE, developed in 3GPP (Third Generation Partnership Project) and providingtrue 4G broadband mobile access, starting from the first version in release 8 and through the continuingevolution to release 10, the latest version of LTE. Release 10, also known as LTE-Advanced,is of particular interest as it is the major technology approved by the ITU as fulfilling the IMTAdvancedrequirements. The description in this book is based on LTE release 10 and thus provides acomplete description of the LTE-Advanced radio access from the bottom up.Chapter 1 gives the background to LTE and its evolution, looking also at the different standardsbodies and organizations involved in the process of defining 4G. It also gives a discussion of the reasonsand driving forces behind the evolution.Chapters 2–6 provide a deeper insight into some of the technologies that are part of LTE and itsevolution. Because of its generic nature, these chapters can be used as a background not only for LTEas described in this book, but also for readers who want to understand the technology behind othersystems, such as WCDMA/HSPA, WiMAX, and CDMA2000.Chapters 7–17 constitute the main part of the book. As a start, an introductory technical overviewof LTE is given, where the most important technology components are introduced based onthe generic technologies described in previous chapters. The following chapters provide a detaileddescription of the protocol structure, the downlink and uplink transmission schemes, and the associatedmechanisms for scheduling, retransmission and interference handling. Broadcast operation andrelaying are also described. This is followed by a discussion of the spectrum flexibility and the associated
標簽：
4g
移動通信
上傳時間：
2022-07-08
上傳用戶：
