beyondthescreen.wordpress.com
Top New Educational Web Tools for 2008-09 | Beyond The Screen
https://beyondthescreen.wordpress.com/2009/01/05/top-new-educational-web-tools-for-2008-09
On teaching, learning, technology, games and virtual worlds. Laquo; Second Life Educational Environments. Web 20 and Education. Top New Educational Web Tools for 2008-09. January 5, 2009 by Don Margulis. It’s difficult to keep on top of all the new technology tools that may be useful for both educational and personal purposes, but fortunately there are some bloggers to assist us in weeding through the possibilities. The first list was compiled by Jane Hart. Center for Learning and Performance Technologies.
mli7.wordpress.com
漫谈在线学习:阴阳论☯ | Large Learning
https://mli7.wordpress.com/2011/04/10/online_learning_3
04/10/2011 in machine learning. 我们将需要最小化的目标函数称之为primal problem. 其对应就有dual problem. 例如常见的 Lagrange Duality. 且在常见情况下,最小化primal等价于最大化dual. 有时候直接求解primal problem很麻烦,但dual却方便很多。 块,从而不能直接对每块做梯度下降得到online gradient descent. 于是我们转向dual problem. Shai 毕业于耶路撒冷希伯来大学,PHD论文便是online learning. 然后他去了online learning重地TTIC, 更是习得满身武艺,现在又回了耶路撒冷希伯来大学任教。 前一种表示是函数的primal form, 而后者则被称之为 Fenchel conjudate. 是一种dual form, 其定义为. 此定理可以为Boosting, learning theory里的Rademecher Bound提供简单的证明,不过我们先关注它在online上的应用。 具体可见Shai的 NIPS ’08. 04/15/20...
mli7.wordpress.com
高效的Matlab:mex & profile | Large Learning
https://mli7.wordpress.com/2011/04/17/efficient_matlab_2
高效的Matlab mex and profile. 这时的一个选择就是混合编程: mex, 它可以实现matlab与C/C /Fortan之间的相互调用。 而matlab却会不谙风情的先将这个大矩阵在内存里复制一次,然后再逐一算exp. 有时我们明明算好内存能存下,但matlab却报out of memory了。 Columbia College Onl…. Mu on 漫谈在线学习 专家问题. TransMatrix on 高效的Matlab mex and profile. Zhaoyin on 漫谈在线学习 专家问题. 漫谈在线学习 阴阳论 增强视觉 …. Wurenzhi on 漫谈在线学习 专家问题. Rospink on 高效的Matlab 向量化. 漫谈在线学习 在线梯度下降 增强视觉…. Tiantian on 漫谈在线学习 阴阳论. Xnature on 漫谈在线学习 专家问题. Mu on 漫谈在线学习 阴阳论. 高效的Matlab mex and profile. Comments feed for this article. 04/18/2011 at 12:00 pm.
beyondthescreen.wordpress.com
“As the Worlds Turn” VWBPE Conference Slides | Beyond The Screen
https://beyondthescreen.wordpress.com/2010/03/14/as-the-worlds-turn-vwbpe-conference-slides
On teaching, learning, technology, games and virtual worlds. Laquo; A New Blog Location. 8220;As the Worlds Turn” VWBPE Conference Slides. March 14, 2010 by Don Margulis. As requested, here are the PowerPoint slides for our two sessions for those attending the Second Life conference on virtual worlds best educational practice:. As the Worlds Turn3-black[1]. The Reality of Transcending the Virtual – SL Conference – Dona Cady 3-13-10. Posted in Teaching and Learning. Tagged virtual worlds education.
mli7.wordpress.com
April | 2011 | Large Learning
https://mli7.wordpress.com/2011/04
You are currently browsing the monthly archive for April 2011. 高效的Matlab mex and profile. 这时的一个选择就是混合编程: mex, 它可以实现matlab与C/C /Fortan之间的相互调用。 而matlab却会不谙风情的先将这个大矩阵在内存里复制一次,然后再逐一算exp. 有时我们明明算好内存能存下,但matlab却报out of memory了。 前面一直在谈理论,为了避免成为纯理论blog, 这次聊聊最实验的部分 coding. For i=1:length(a), a(i)=a(i) 1; end. 虽然matlab在不断优化for的执行,不过即使是matlab 2010b, 在我的笔记本上 Core 2 Duo 2.4 GHz, 4GB, Win7 ,对于一个10万元素的一维数组,第一种实现需5微秒,而第二种只用0.4微妙。 例如all 是否元素全非0 , cumsum 累加 , diff 相邻元素的差 ,prod 累乘。 但如果我们想用”外加”呢 既. 我们将需要最小化的目标函数称之为primal pr...
mli7.wordpress.com
March | 2011 | Large Learning
https://mli7.wordpress.com/2011/03
You are currently browsing the monthly archive for March 2011. 03/25/2011 in machine learning. 它是对在线学习 online learning 最贴切的刻画 数据不断前来,我们需根据当前所能得到的来调整自己的最优策略。 熟悉机器学习的稳拿可能注意到了在线学习与在机器学习 统计学习 里所讨论的 离线 学习的区别。 而向来被优化界抛弃的在线学习、随机优化算法 例如stochastic gradient descent ,虽然收敛不快,不过迭代代价很低,更考虑到learning算法的解的精度要求不高,所以在实际应用中这些方法通常比传统的优化算法快很多,而且可以处理非常大的数据。 一个通常的规律是learning界火过的topic在computer vision,data mining等偏应用的方向上会接着火几年,所以肯定CVPR, ICCV, KDD之类会议上还会有茫茫多这类paper。 Columbia College Onl…. Mu on 漫谈在线学习 专家问题. Zhaoyin on 漫谈在线学习 专家问题.
beyondthescreen.wordpress.com
Second Life Educational Environments | Beyond The Screen
https://beyondthescreen.wordpress.com/2008/04/05/second-life-educational-environments
On teaching, learning, technology, games and virtual worlds. Laquo; Virtual World Academic Resources. Top New Educational Web Tools for 2008-09. Second Life Educational Environments. April 5, 2008 by Don Margulis. Is an open ended virtual world that is popular with educators interested in the potential of this technology. The environments in. There are three ways to travel to a specific location:. A The first method is to click on one of the links below. Here are some sites to consider visiting:. This is...
yourcareerinnursing.com
RT to RN Online Bridge Program
http://www.yourcareerinnursing.com/rt-to-rn-bridge-program
Your Career In Nursing. RT to RN Bridge Programs Employment. RT to RN Program. Starting a Healthcare Career. RT to RN Online Bridge Program. Advance Your Career from RT to RN Online. Earn your degree from an accredited college and obtain your RN license, then enjoy increased stability, prestige, and salary benefits. 10 Reasons to Bridge from RT to RN. No classes to attend. Convenient programs that fit into your schedule. Save money over traditional programs. High quality programs with guaranteed financing.
mli7.wordpress.com
高效的Matlab:向量化 | Large Learning
https://mli7.wordpress.com/2011/04/17/efficient_matlab
前面一直在谈理论,为了避免成为纯理论blog, 这次聊聊最实验的部分 coding. For i=1:length(a), a(i)=a(i) 1; end. 虽然matlab在不断优化for的执行,不过即使是matlab 2010b, 在我的笔记本上 Core 2 Duo 2.4 GHz, 4GB, Win7 ,对于一个10万元素的一维数组,第一种实现需5微秒,而第二种只用0.4微妙。 例如all 是否元素全非0 , cumsum 累加 , diff 相邻元素的差 ,prod 累乘。 Bsxfun是一个matlab自版本R2007a来就提供的一个函数,作用是”applies an element-by-element binary operation to arrays a and b, with singleton expansion enabled.”. A = randn(3,1), b = randn(1,3) a = -0.2453 -0.2766 -0.1913 b = 0.6062 0.5655 0.9057. 但如果我们想用”外加”呢 既. Mu on 漫谈在线学习 专家问题.
SOCIAL ENGAGEMENT