MLOps is about taking experimental ML models to production, i.e., serving the models to actual users. Unfortunately, existing ML serving systems do not adequately handle the dynamic environments in which online data diverges from offline training dat
a, resulting in tedious model updating and deployment works. This paper implements a lightweight MLOps plugin, termed ModelCI-e (continuous integration and evolution), to address the issue. Specifically, it embraces continual learning (CL) and ML deployment techniques, providing end-to-end supports for model updating and validation without serving engine customization. ModelCI-e includes 1) a model factory that allows CL researchers to prototype and benchmark CL models with ease, 2) a CL backend to automate and orchestrate the model updating efficiently, and 3) a web interface for an ML team to manage CL service collaboratively. Our preliminary results demonstrate the usability of ModelCI-e, and indicate that eliminating the interference between model updating and inference workloads is crucial for higher system efficiency.
To mitigate the spread of COVID-19 pandemic, decision-makers and public authorities have announced various non-pharmaceutical policies. Analyzing the causal impact of these policies in reducing the spread of COVID-19 is important for future policy-ma
king. The main challenge here is the existence of unobserved confounders (e.g., vigilance of residents). Besides, as the confounders may be time-varying during COVID-19 (e.g., vigilance of residents changes in the course of the pandemic), it is even more difficult to capture them. In this paper, we study the problem of assessing the causal effects of different COVID-19 related policies on the outbreak dynamics in different counties at any given time period. To this end, we integrate data about different COVID-19 related policies (treatment) and outbreak dynamics (outcome) for different United States counties over time and analyze them with respect to variables that can infer the confounders, including the covariates of different counties, their relational information and historical information. Based on these data, we develop a neural network based causal effect estimation framework which leverages above information in observational data and learns the representations of time-varying (unobserved) confounders. In this way, it enables us to quantify the causal impact of policies at different granularities, ranging from a category of policies with a certain goal to a specific policy type in this category. Besides, experimental results also indicate the effectiveness of our proposed framework in capturing the confounders for quantifying the causal impact of different policies. More specifically, compared with several baseline methods, our framework captures the outbreak dynamics more accurately, and our assessment of policies is more consistent with existing epidemiological studies of COVID-19.
DNN-based video analytics have empowered many new applications (e.g., automated retail). Meanwhile, the proliferation of fog devices provides developers with more design options to improve performance and save cost. To the best of our knowledge, this
paper presents the first serverless system that takes full advantage of the client-fog-cloud synergy to better serve the DNN-based video analytics. Specifically, the system aims to achieve two goals: 1) Provide the optimal analytics results under the constraints of lower bandwidth usage and shorter round-trip time (RTT) by judiciously managing the computational and bandwidth resources deployed in the client, fog, and cloud environment. 2) Free developers from tedious administration and operation tasks, including DNN deployment, cloud and fogs resource management. To this end, we implement a holistic cloud-fog system referred to as VPaaS (Video-Platform-as-a-Service). VPaaS adopts serverless computing to enable developers to build a video analytics pipeline by simply programming a set of functions (e.g., model inference), which are then orchestrated to process videos through carefully designed modules. To save bandwidth and reduce RTT, VPaaS provides a new video streaming protocol that only sends low-quality video to the cloud. The state-of-the-art (SOTA) DNNs deployed at the cloud can identify regions of video frames that need further processing at the fog ends. At the fog ends, misidentified labels in these regions can be corrected using a light-weight DNN model. To address the data drift issues, we incorporate limited human feedback into the system to verify the results and adopt incremental learning to improve our system continuously. The evaluation demonstrates that VPaaS is superior to several SOTA systems: it maintains high accuracy while reducing bandwidth usage by up to 21%, RTT by up to 62.5%, and cloud monetary cost by up to 50%.
Deep learning (DL) models have become core modules for many applications. However, deploying these models without careful performance benchmarking that considers both hardware and softwares impact often leads to poor service and costly operational ex
penditure. To facilitate DL models deployment, we implement an automatic and comprehensive benchmark system for DL developers. To accomplish benchmark-related tasks, the developers only need to prepare a configuration file consisting of a few lines of code. Our system, deployed to a leader server in DL clusters, will dispatch users benchmark jobs to follower workers. Next, the corresponding requests, workload, and even models can be generated automatically by the system to conduct DL serving benchmarks. Finally, developers can leverage many analysis tools and models in our system to gain insights into the trade-offs of different system configurations. In addition, a two-tier scheduler is incorporated to avoid unnecessary interference and improve average job compilation time by up to 1.43x (equivalent of 30% reduction). Our system design follows the best practice in DL clusters operations to expedite day-to-day DL service evaluation efforts by the developers. We conduct many benchmark experiments to provide in-depth and comprehensive evaluations. We believe these results are of great values as guidelines for DL service configuration and resource allocation.
MLModelCI provides multimedia researchers and developers with a one-stop platform for efficient machine learning (ML) services. The system leverages DevOps techniques to optimize, test, and manage models. It also containerizes and deploys these optim
ized and validated models as cloud services (MLaaS). In its essence, MLModelCI serves as a housekeeper to help users publish models. The models are first automatically converted to optimized formats for production purpose and then profiled under different settings (e.g., batch size and hardware). The profiling information can be used as guidelines for balancing the trade-off between performance and cost of MLaaS. Finally, the system dockerizes the models for ease of deployment to cloud environments. A key feature of MLModelCI is the implementation of a controller, which allows elastic evaluation which only utilizes idle workers while maintaining online service quality. Our system bridges the gap between current ML training and serving systems and thus free developers from manual and tedious work often associated with service deployment. We release the platform as an open-source project on GitHub under Apache 2.0 license, with the aim that it will facilitate and streamline more large-scale ML applications and research projects.
In this paper, we first present simple proofs of Chois results [4], then we give a short alternative proof for Fiedler and Markhams inequality [6]. We also obtain additional matrix inequalities related to partial determinants.
Let $(mathbb M, d,mu)$ be a metric measure space with upper and lower densities: $$ begin{cases} |||mu|||_{beta}:=sup_{(x,r)in mathbb Mtimes(0,infty)} mu(B(x,r))r^{-beta}<infty; |||mu|||_{beta^{star}}:=inf_{(x,r)in mathbb Mtimes(0,infty)} mu(B(x,r)
)r^{-beta^{star}}>0, end{cases} $$ where $beta, beta^{star}$ are two positive constants which are less than or equal to the Hausdorff dimension of $mathbb M$. Assume that $p_t(cdot,cdot)$ is a heat kernel on $mathbb M$ satisfying Gaussian upper estimates and $mathcal L$ is the generator of the semigroup associated with $p_t(cdot,cdot)$. In this paper, via a method independent of Fourier transform, we establish the decay estimates for the kernels of the fractional heat semigroup ${e^{-t mathcal{L}^{alpha}}}_{t>0}$ and the operators ${{mathcal{L}}^{theta/2} e^{-t mathcal{L}^{alpha}}}_{t>0}$, respectively. By these estimates, we obtain the regularity for the Cauchy problem of the fractional dissipative equation associated with $mathcal L$ on $(mathbb M, d,mu)$. Moreover, based on the geometric-measure-theoretic analysis of a new $L^p$-type capacity defined in $mathbb{M}times(0,infty)$, we also characterize a nonnegative Randon measure $ u$ on $mathbb Mtimes(0,infty)$ such that $R_alpha L^p(mathbb M)subseteq L^q(mathbb Mtimes(0,infty), u)$ under $(alpha,p,q)in (0,1)times(1,infty)times(1,infty)$, where $u=R_alpha f$ is the weak solution of the fractional diffusion equation $(partial_t+ mathcal{L}^alpha)u(t,x)=0$ in $mathbb Mtimes(0,infty)$ subject to $u(0,x)=f(x)$ in $mathbb M$.
Let $mathcal{H}_{alpha}=Delta-(alpha-1)|x|^{alpha}$ be an $[1,infty) ialpha$-Hermite operator for the hydrogen atom located at the origin in $mathbb R^d$. In this paper, we are motivated by the classical case $alpha=1$ to investigate the space of fun
ctions with $alpha$-{it Hermite Bounded Variation} and its functional capacity and geometrical perimeter.
In this paper, a compact and low-cost structured illumination microscope (SIM) based on a 2X2 fiber coupler is presented. Fringe illumination is achieved by placing two output fiber tips at a conjugate Fourier plane of the sample plane as the point s
ources. Raw structured illumination (SI) images in different pattern orientations are captured when rotating the fiber mount. Following this, high resolution images are reconstructed from no-phase-shift raw SI images by using a joint Richardson-Lucy (jRL) deconvolution algorithm. Compared with an SLM-based SIM system, our method provides a much shorter illumination path, high power efficiency, and low cost.
The optical interferometry has been widely used in various high precision applications. Usually, the minimum precision of an interferometry is limited by various technique noises in practice. To suppress such kind of noises, we propose a novel scheme
, which combines the weak measurement with the standard interferometry. The proposed scheme dramatically outperforms the standard interferometry in the signal noise ratio and the robustness against noises caused by the optical elements reflections and the offset fluctuation between two paths. A proof-of-principle experiment is demonstrated to validate the amplification theory.
