Lecture Notes on Support Vector Machine Feng Li fli@sdu.edu.cn Shandong University, China 1 Hyperplane and Margin In a n-dimensional space, a hyper plane is defined by ωT x + b = 0 (1) where ω ∈ Rn the margin is defined as γ = min i γ(i) (6) 1 ? ? ! ? ! Figure 1: Margin and hyperplane. 2 Support Vector Machine 2.1 Formulation The hyperplane actually serves as a decision boundary to differentiating samples are so-called support vector, i.e., the vectors “supporting” the margin boundaries. We can redefine ω by w = � s∈S αsy(s)x(s) where S denotes the set of the indices of the support vectors 4 Kernel

Lecture 6: Support Vector Machine Feng Li Shandong University fli@sdu.edu.cn December 28, 2021 Feng Li (SDU) SVM December 28, 2021 1 / 82 Outline 1 SVM: A Primal Form 2 Convex Optimization Review parallely along ω (b < 0 means in opposite direction) Feng Li (SDU) SVM December 28, 2021 3 / 82 Support Vector Machine A hyperplane based linear classifier defined by ω and b Prediction rule: y = sign(ωTx Scaling ! and " such that min& ' & !() & + " = 1 Feng Li (SDU) SVM December 28, 2021 14 / 82 Support Vector Machine (Primal Form) Maximizing 1/∥ω∥ is equivalent to minimizing ∥ω∥2 = ωTω min ω,b ωTω

SberBank story: moving Istio from PoC to production Igor Gustomyasov, Sber Maksim Chudnovskii, IBM Sber position across key areas Best client experience Technological leadership In financial services ExportTo tuning is required 1. Resource consumption 2. Resource Mounts (#15517) 4. Tests on the production-size environment aren’t a waste of time 1. Istio Discovery Restarts (#25495) 2. Proxy Probes (#26792) Discovery for OCP & Kubernetes • Multi-cluster Service Topology • Cloud-Native Event Hub • Full Support for VM-Based Workloads • UX Simplification CONTACT US Head of integration department Igor Gustomyasov

Kubernetes for Edge Computing across Inter-Continental Haier Production Sites Jiyuan Tang & Xin Zhang zhangxin@caicloud.io tangjiyuan@caicloud.io 关于我们 • 开源技术创新者 • 从 Kubernetes 到 Kubeflow • Google

predicted that more than 75% of worldwide organizations would run containerized applications in production by 2022. Confidence of this prediction demonstrated the value of cloud native technologies for and security. Relying on upstream Kubernetes isn't enough for teams deploying Kubernetes into production. Basic Kubernetes installations are plagued by a lack of central visibility, inconsistent security Managed Services 3 3 1 2 Multi-Cluster Management 4 3 2 2 Edge Support 4 3 2 1 Integrated Public Cloud Support 4 2 2 2 Bare Metal, OpenStack & vSphere 4 3 2 2 Import Existing

................................................................ 9 How SUSE Rancher delivers production-grade Kubernetes at scale: ..................................... 10 Solution architecture .. giving the DevOps team freedom to build and run containerized applications anywhere. The PowerFlex family offers key value propositions for traditional and cloud-native production workloads, deployment PowerProtect Data Manager is used to discover, protect, and restore production workloads in Kubernetes environments and protects production and development, or test workloads to ensure that the data is easy

customizations to fit it into different envi- ronments, but it’s difficult to say which is a hardened, production-ready approach. Having a secured profile with an opinionated cluster configuration will help guide Route Hijacking 023 High Pilot Debug Interface Exposes Sensitive Information 002 Medium Default Production Profile Not Sufficiently Hardened 003 Medium Weak Hash Used for Integrity 009 Medium Go Trace between the control plane will be secure by default.”1 In the “Default” profile used to represent a production environment, the “controlPlaneAuthP olicy” is set to “NONE” instead of “mTLS”: mesh: |- ... defaultConfig:

terms of resources, the project was assigned to four members of the Cure53 team with best-suited expertise and skills. The testing team examined the scope in June 2020, namely in calendar weeks 24 and 25. A during the project stood at twenty person-days. To best address the objectives expressed by the Dapr team, two work packages (WPs) were outlined. In WP1, Cure53 performed both a broad and thorough source shift in methods, WP2 encompassed penetration tests against Dapr integration and setup. The Cure53 team relied on a fully installed Kubernetes cluster, complete with sample applications that needed to be

meetings with the Istio team to discuss questions and issues that came out throughout the period of the audit. Found issues were reported as they came up which gave the Istio team time to triage and assess investigation that revealed a vulnerability in Golang itself. The finding was reported by the auditing team to the Istio maintainers, because Istio does not cap the size of requests made on an h2c connection managed Istio offering which has MultiplexHTTP configured. A�er issue 10 had been reported to the Istio team, Istio maintainer John Howard assessed Golangs recommended solution for capping H2c requests which

Brandon Gul- la, Tim Nicklas, Chris Nuber and the team at SUSE Rancher Government Services (SUSE RGS, formerly Rancher Federal), the Booz Allen team is underpinning its smart, tactical edge platform Kubernetes distribution, designed for production workloads in unat- tended, resource constrained, remote envi- ronments. This is the story of how, alongside the team at SUSE RGS, Booz Allen is delivering SmartEdge infrastruc- ture, containers and Kubernetes were part of the plan from the start. The team began experimenting with Ku- bernetes and early versions of Docker in 2015. They soon became familiar