Deep learning is a form of machine learning that enables computers to learn from experience and understand the world in terms of a hierarchy of concepts. Because the computer gathers knowledge from experience, there is no need for a human computer operator to formally specify all the knowledge that the computer needs. The hierarchy of concepts allows the computer to learn complicated concepts by building them out of simpler ones; a graph of these hierarchies would be many layers deep. This book introduces a broad range of topics in deep learning. The text offers mathematical and conceptual background, covering relevant concepts in linear algebra, probability theory and information theory, numerical computation, and machine learning. It describes deep learning techniques used by practitioners in industry, including deep feedforward networks, regularization, optimization algorithms, convolutional networks, sequence modeling, and practical methodology; and it surveys such applications as natural language processing, speech recognition, computer vision, online recommendation systems, bioinformatics, and videogames. Finally, the book offers research perspectives, covering such theoretical topics as linear factor models, autoencoders, representation learning, structured probabilistic models, Monte Carlo methods, the partition function, approximate inference, and deep generative models. Deep Learning can be used by undergraduate or graduate students planning careers in either industry or research, and by software engineers who want to begin using deep learning in their products or platforms. A website offers supplementary material for both readers and instructors.

Deep Learning

We investigate conditional adversarial networks as a general-purpose solution to image-to-image translation problems. These networks not only learn the mapping from input image to output image, but also learn a loss function to train this mapping. This makes it possible to apply the same generic approach to problems that traditionally would require very different loss formulations. We demonstrate that this approach is effective at synthesizing photos from label maps, reconstructing objects from edge maps, and colorizing images, among other tasks. Moreover, since the release of the pix2pix software associated with this paper, hundreds of twitter users have posted their own artistic experiments using our system. As a community, we no longer hand-engineer our mapping functions, and this work suggests we can achieve reasonable results without handengineering our loss functions either.

/pdf/image-to-image-translation-with-conditional-adversarial-5058t24et2.pdf

Image-to-Image Translation with Conditional Adversarial Networks

Image-to-image translation is a class of vision and graphics problems where the goal is to learn the mapping between an input image and an output image using a training set of aligned image pairs. However, for many tasks, paired training data will not be available. We present an approach for learning to translate an image from a source domain X to a target domain Y in the absence of paired examples. Our goal is to learn a mapping G : X → Y such that the distribution of images from G(X) is indistinguishable from the distribution Y using an adversarial loss. Because this mapping is highly under-constrained, we couple it with an inverse mapping F : Y → X and introduce a cycle consistency loss to push F(G(X)) ≈ X (and vice versa). Qualitative results are presented on several tasks where paired training data does not exist, including collection style transfer, object transfiguration, season transfer, photo enhancement, etc. Quantitative comparisons against several prior methods demonstrate the superiority of our approach.

Unpaired Image-to-Image Translation Using Cycle-Consistent Adversarial Networks

We investigate conditional adversarial networks as a general-purpose solution to image-to-image translation problems. These networks not only learn the mapping from input image to output image, but also learn a loss function to train this mapping. This makes it possible to apply the same generic approach to problems that traditionally would require very different loss formulations. We demonstrate that this approach is effective at synthesizing photos from label maps, reconstructing objects from edge maps, and colorizing images, among other tasks. Indeed, since the release of the pix2pix software associated with this paper, a large number of internet users (many of them artists) have posted their own experiments with our system, further demonstrating its wide applicability and ease of adoption without the need for parameter tweaking. As a community, we no longer hand-engineer our mapping functions, and this work suggests we can achieve reasonable results without hand-engineering our loss functions either.

TensorFlow is a machine learning system that operates at large scale and in heterogeneous environments. Tensor-Flow uses dataflow graphs to represent computation, shared state, and the operations that mutate that state. It maps the nodes of a dataflow graph across many machines in a cluster, and within a machine across multiple computational devices, including multicore CPUs, general-purpose GPUs, and custom-designed ASICs known as Tensor Processing Units (TPUs). This architecture gives flexibility to the application developer: whereas in previous "parameter server" designs the management of shared state is built into the system, TensorFlow enables developers to experiment with novel optimizations and training algorithms. TensorFlow supports a variety of applications, with a focus on training and inference on deep neural networks. Several Google services use TensorFlow in production, we have released it as an open-source project, and it has become widely used for machine learning research. In this paper, we describe the TensorFlow dataflow model and demonstrate the compelling performance that TensorFlow achieves for several real-world applications.

/pdf/tensorflow-a-system-for-large-scale-machine-learning-10wgdjhl6k.pdf

TensorFlow: a system for large-scale machine learning

Sinomenine (SIN) has long been known as an anti-inflammatory drug, while poor efficiency and large-dose treatment had limited its further application. A series of novel SIN derivatives 1–26 were designed and synthesized to improve its anti-inflammatory activity. The anti-inflammatory activity evaluation showed most of the derivatives exhibited enhanced anti-inflammatory activity in vitro compared to SIN. Compound 17 significantly inhibited LPS-induced secretion of pro-inflammatory factors NO (IC50 = 30.28 ± 1.70 μM), and suppressed the expression of iNOS, IL-6 and TNF-α in RAW264.7 cells. Moreover, compound 17 showed excellent anti-inflammatory in mouse paw edema. Immunohistochemistry results revealed that compound 17 exerted anti-inflammatory activity by inhibiting the pro-inflammatory cytokine TNF-α. Furthermore, compound 17 exhibited an analgesic effect in vivo. The results attained in this study indicated that compound 17 had the potential to be developed into an anti-inflammation and analgesic agent.

/pdf/synthesis-and-biological-evaluation-of-novel-sinomenine-qsfwy1u1.pdf

Synthesis and biological evaluation of novel sinomenine derivatives as anti-inflammatory and analgesic agent

Podophyllotoxin's undifferentiated cytotoxicity and poor selectivity limit its clinical application. To improve above disadvantages, conjugation of bile acids with podophyllotoxin could improve cell line selectivity of liver cancer to achieve clinical translation further. Enlightened by the bile acids' moiety magic characters, thirty podophyllotoxin-linked bile acid derivatives had been designed and synthesized. The cytotoxicity of these compounds in vitro was evaluated on HepG2, HCT-116, A549 and MDCK cell lines. After conjunction with bile acids, most of the derivatives (IC50 = 0.066-0.831 μM) were more potent against above three types of tumor cells than Etoposide (VP-16, IC50 = 4.319-41.080 μM) and exhibited similar antitumor activity compared with doxorubicin (DOX, IC50 = 0.230-0.745 μM). Moreover, structure-activity relationship displayed the length of the linker chain between podophyllotoxin and bile acids affected the cytotoxicity. Especially, compound 23 exhibited strong activity against HepG2 cell lines (IC50 = 0.188 ± 0.01 μM) than MDCK cell lines (IC50 = 4.780 ± 0.50 μM) and its SI (IC50MDCK/IC50HepG2) value of compound 23 was 25.4. Further antitumor mechanism studies showed that compound 23 acted as Topo Ⅱ inhibition and induced cell apoptosis with S cell cycle arrest. In particular, compound 23 showed valid antitumor efficacy at 10 mg/kg by intraperitoneal administration with a tumor inhibition rate of 60.9% in the Hepa1-6 xenograft mice model. The current research displayed that introduction of bile acids contributed to improve selectivity and activity to cell, and compound 23 could be a promising anti-tumor candidate.

Synthesis and biological activity evaluation of podophyllotoxin-linked bile acid derivatives as potential anti-liver cancer agents.

Design and Synthesis of Novel Anti-Multidrug-Resistant Staphylococcus Aureus Derivatives of Glycyrrhetinic Acid by Blocking Arginine Biosynthesis, Metabolic and H 2S Biogenesis

Explore bioactive ingredients and potential mechanism of Houpo Mahuang decoction for chronic bronchitis based on UHPLC-Q exactive orbitrap HRMS, network pharmacology, and experiment verification.

Bing Xu

Papers

Synthesis and biological evaluation of novel sinomenine derivatives as anti-inflammatory and analgesic agent

Synthesis and biological activity evaluation of podophyllotoxin-linked bile acid derivatives as potential anti-liver cancer agents.

Design and Synthesis of Novel Anti-Multidrug-Resistant Staphylococcus Aureus Derivatives of Glycyrrhetinic Acid by Blocking Arginine Biosynthesis, Metabolic and H 2S Biogenesis

Explore bioactive ingredients and potential mechanism of Houpo Mahuang decoction for chronic bronchitis based on UHPLC-Q exactive orbitrap HRMS, network pharmacology, and experiment verification.

Mechanism of Anti-Ad Effect of Kai-Xin-San Based on UPLC-Q-Exactive Ms, Network Pharmacology and Experimental Verification