Deep convolutional neural networks (CNNs) have achieved breakthrough performance in many pattern recognition tasks such as image classification. However, the development of high-quality deep models typically relies on a substantial amount of trial-and-error, as there is still no clear understanding of when and why a deep model works. In this paper, we present a visual analytics approach for better understanding, diagnosing, and refining deep CNNs. We formulate a deep CNN as a directed acyclic graph. Based on this formulation, a hybrid visualization is developed to disclose the multiple facets of each neuron and the interactions between them. In particular, we introduce a hierarchical rectangle packing algorithm and a matrix reordering algorithm to show the derived features of a neuron cluster. We also propose a biclustering-based edge bundling method to reduce visual clutter caused by a large number of connections between neurons. We evaluated our method on a set of CNNs and the results are generally favorable.

Interactive model analysis, the process of understanding, diagnosing, and refining a machine learning model with the help of interactive visualization, is very important for users to efficiently solve real-world artificial intelligence and data mining problems. Dramatic advances in big data analytics has led to a wide variety of interactive model analysis tasks. In this paper, we present a comprehensive analysis and interpretation of this rapidly developing area. Specifically, we classify the relevant work into three categories: understanding, diagnosis, and refinement. Each category is exemplified by recent influential work. Possible future research opportunities are also explored and discussed. Keywords: interactive model analysis, interactive visualization, machine learning, understanding, diagnosis, refinement IntroductionMachine learning has been successfully applied to a wide variety of fields ranging from information retrieval, data mining, and speech recognition, to computer graphics, visualization, and human-computer interaction. However, most users often treat a machine learning model as a black box because of its incomprehensible functions and unclear working mechanism [1,2,3]. Without a clear understanding of how and why a model works, the development of highperformance models typically relies on a time-consuming trial-and-error pro-$ Fully documented templates are available in the elsarticle package on CTAN.

A new and efficient approach using cleaving of trimethylsilyl groups to create covalent Au-C anchoring sites has been developed for single-molecule junction conductance measurements. Employing the mechanically controllable break junction (MCBJ) technique in liquid, we demonstrate the formation of highly conducting single molecular junctions of several OPE derivatives. The created junctions are mechanically stable and exhibit conductances around one order of magnitude higher than those of their dithiol analogues. Extended assembly and reaction times lead to oligomerization. Combined STM imaging and gap-mode Raman experiments provide structure evidence to support the formation of covalent Au-C contacts and further oligomerization.

To study the electronic interactions in donor–acceptor (D–A) ensembles, D and A fragments are coupled in a single molecule. Specifically, a tetrathiafulvalene (TTF)‐fused dipyrido[3,2‐a:2′,3′‐c]phenazine (dppz) compound having inherent redox centers has been synthesized and structurally characterized. Its electronic absorption, fluorescence emission, photoinduced intramolecular charge transfer, and electrochemical behavior have been investigated. The observed electronic properties are explained on the basis of density functional theory.