Friday, December 28, 2018

Golden-Nugget Shares - Business, CRE, Tech, Blockchain and AI - December 17-28

Also publish on LinkedIn and our GoogleGroup InfoPortal


We use a powerful method of sharing - post “Golden Nuggets of Information” on Social Media multiple times daily or in regular emails (every 5-15 days).  We’ve been sharing the essence of articles and a few powerful comments regularly.* 

Commercial Real Estate, Technology, Blockchain and AI

Kirk Borne.  COLD-START PROBLEM. "How does #DataScience #Model-building get started towards the optimal solution? Make a guess! Then work towards increasing performance using an #ObjectiveFunction - cost or benefit - a measure of model performance. Make incremental adjustments, correcting towards the better perfomance." (#Technology #AI #MachineLearning #DeepLearning #Analysis #Business #Research #Expansion; heavily edited; examples follow - via
  • "#CLUSTERING ANALYSIS (such as K-Means Clustering), where the initial cluster means and counts are not known in advance (thus initialized randomly), but the compactness of the clusters can be used to evaluate, iterate and improve the set of clusters in a progression to the final optimum set of clusters."
  • "#NEURAL NETWORKS, where the initial weights on the network edges are assigned randomly (a cold-start), but #backpropagation is used to iterate the model to the optimal network (with highest classification performance)." 
  • "#TENSORFLOW #DeepLearning, which uses the same #backpropagation technique of simpler #neural networks, but the calculation of the weight adjustments is made across a very high-dimensional parameter space of deep network layers and edge weights using tensors."
  • "#REGRESSION, which uses the sum of the squares of the deviations of the points from the model curve in order to find the best-fit curve. Linear regression is well defined.  Non-linear regression is not typically a closed-form set of mathematical equations [and requires a guess at its form], but the minimization of the sum of the squares of deviations still applies."
  • "#NONCONVEX OPTIMIZATION, where the objective function has many hills and valleys, so that gradient descent and hill-climbing will typically converge only to a local optimum, not to the global optimum. There are techniques to generate lots of random (cold-start) models and then iterate each of them until you find the global optimum.  See his graphic and note on genetic algorithms - these are not machine learning algorithms specifically, but they are actually meta-learning algorithms."
  • "NAIVE #BAYES CLASSIFICATION, which applies Bayes probability theorem to a large data set for which some combinations of attributes and features are not represented in the training data - a cold-start challenge.  Assuming different attributes are mutually independent, then one can estimate the posterior likelihood for what the class label should be for a new data item with a feature vector (set of attributes) that is not found in the training data." 
  • "SOCIAL NETWORK (link) ANALYSIS, where the patterns in the network - centrality, reach, degrees of separation, density, cliques, etc - encode knowledge about the network, such as the most authoritative or influential nodes in the network.  Applies algorithms like PageRank, without any prior knowledge about those patterns - a cold-start."  Also see #kNN, #MARKOV MODELING and ASSOCIATION RULE MINING.
Hassan Syed noted Guardian News & Media:  BLOCKCHAIN - SO MUCH BIGGER than BITCOIN - "may just change the way we run our lives, since BC allows the authentication of transactions without them needing to be administered or guaranteed by a central authority (#Technology #Analysis #Business #Data #Research #Investment #Startup; edited
  • VOTING. Ballot boxes and current online voting platforms are vulnerable to manipulation; now a startup called Follow My Vote is developing a blockchain-based system to ensure #security, #transparency and mathematically accurate election #results."
  • "#POWER GENERATION.  BC enables homeowners to sell back energy to the grid bypassing the energy provider, or manage their own, independent microgrids. The BC allows them to set their own price, without an price-setting, commission-taking intermediary."
  • "#MUSIC #STREAMING.  Prepare to have your listening habits disrupted by BC-based music streaming services, without a service such as Apple Music or Spotify taking a cut.  A BC system called Voise is enabling artists to set a price of which they receive 100% of user-streamed music."
  • "#HEALTHCARE.  Medical records are often scattered between GPs, clinics and labs. A BC-based health record could be read and updated from multiple locations or services and would contain the adders' IDs. The patient can choose whom to share it with. Example - MIT's MedRec."  It will take strong effort and "an Act of Congress" to be approved within #HIPPA law."
  • #AID.  In 2012, the UN estimated that 30% of development aid was lost to corruption; they have BC-based projects looking to solve issues in delivering aid.  In a food programme pilot project, refugees made purchases at the camp supermarket, authenticated by iris scans, and their spending deducted from their #Crypto allowance, cutting transaction fees and the frequency of fraud and theft."  Hopefully, the original food distribution will be tracked by BC to make this work well.

John Carpenter noted Hyperledger. "#HYPERLEDGER URSA is a modular, flexible #blockchain #cryptography #library to make it much safer and easier for our #DistributedLedger projects to use existing, time tested, and trusted cryptographic libraries but also new cryptographic library implementations. It is much better to collaborate on a shared library for many reasons, including: avoiding duplication, security, expert review, cross-platform interoperability, modularity and new project launches." (#Technology #AI #Business #Data #Research #Expansion #Investment #SmallBusiness #Startup; edited - via
  • "AVOIDING DUPLICATION. #Crypto implementations are notoriously difficult to get correct and often require much work to achieve a high level of #security. The library allows projects to share crypto implementations, avoiding unnecessary duplication and extra work."
  • "SECURITY. Having most of the crypto #code in a single location substantially simplifies the security #analysis of the crypto portion of #Hyperledger. This means maintenance is easier and more efficient. The presence of easy-to-use, secure crypto implementations might also make it less likely for less-experienced people to create less-secure implementations."
  • "Expert REVIEW. The ability to enforce expert review of all #cryptographic #code should increase #security as well. Having all of our cyptographic code in a single location makes it easier to concentrate all of the cryptographic expertise in the project and ensures that code will be well reviewed, thus decreasing the likelihood of dangerous security bugs.
  • "Cross-Platform #INTEROPERABILITY. If two projects use the same crypto libraries, it simplifies cross-platform interoperability, since cryptographic verification involves the same protocols on both sides."
  • "#MODULARITY. This could be the first common component/module and a step towards modular #DistributedLedger platforms, which share common components. This modularity brings in terms of actual functionality, a successful #crypto library encourages and pushes forward more modular activities.
  • "NEW PROJECT LAUNCHES. It is easier for new projects to get off the ground if they have easy access to well-implemented, modular cryptographic abstractions."
  • "FUTURE PLANS: "Focused on a shared modular-#signature #library, but exact implementations and APIs have not been finalized. Beyond that, they expect other wrappings and modular code to go in this library: use aggregate signatures, #hash #algorithms, selecting vetted implementations and providing a common #interface."

Pedro Uria-Recio. MISTAKES CORPORATES MAKE IN THEIR ADVANCED #ANALYTICS PROGRAMS (#Technology #Analysis #Business #Data #Research #Expansion #Startup; edited; previous post
  • "#STRATEGY:  Day-dreaming that analytics is a plug & play magic wand that will bring very short-term #ROI • Solving problems that are not really worth solving • Relying solely on vendors or consultants for analytics, especially on model creation • Not developing a fully comprehensive list of priorities • Saying yes to random requests • Assuming that abstinence from external data monetization or from cloud is the solution to data privacy and security."
  • "PEOPLE. Organizing analytics under functions which do not drive the business on a daily basis • Letting multiple analytics teams flourish with organizational silos among them • Attracting talent only through base compensation • Hiring a bunch of PhDs who strive to develop highly nuanced models instead of directionally correct rough-and-ready solutions • Hiring a technical or non-technical Chief Data Analytics Officer - needs to be both • Not bringing domain experts and internal business consultants to the analytics teams • Neglecting the creation of a data-driven culture through active coaching across the whole organizations • Not being objective enough and remaining biased to the status quo or leadership thinking." 
  • "EXECUTION:  Not embedding analytics in the operating models and day-to-day workflows • Not collocating data scientists with the business teams they support • Managing analytics projects in waterfall • Not being able to scale analytics pilots up • Neglecting data governance as a fundamental enabler."
  • "#TECHNOLOGY:  Trying to create data science models without refining your data engineering infrastructure • Not using any of the following basic technologies • Having technological siloes among data repositories which makes it difficult to integrate different kinds of data into a model • Not automating analytics through A.I., which can be an extremely smart assistant to data scientists."
  • "#FINANCE:  Not allocating enough budget • Not measuring the ROI."

Commercial Real Estate and Business
Don Catalano. TOP DEAL BREAKERS for #CRE TENANTS (#Analysis #Business #Data #Research #Expansion #Investment #Startup; edited -
  • "EXCESSIVELY LONG #TERMS. Having tenants committed for years at a time means not having to go through the hassles and expenses of filling vacant spaces. But for the tenant, who may find that business needs remaining unchanged over the years, it's more likely that what's best tomorrow won't be the same as what's ideal today."
  • "EXCESSIVE / UNCLEAR #FEES, which are a common part of CRE leases, so examine them closely:  What's the going rate for a particular fee in the local market • What are the fees for, when assessed, how much and when due • If you can't pin them down in clear language added to the contract, look elsewhere." 
  • "#FORFEITURE OF RIGHTS.  You should be able to move into a commercial space in good faith, believing the contract terms.  But ... there are unscrupulous landlords and unforeseeable circumstances after you move in.  If a disaster happens, you need an out, so NEVER waive your legal rights to take any type of action against your landlord."
  • "UNWILLINGNESS to PUT THINGS in WRITING.  Put promises in writing - verbal agreements are not enforceable in a court of law - if landlord refuses, restart your search.  Adding a new clause, inserting a sentence or rewording an aspect of a contract only takes a few minutes.
  • "UNFAVORABLE RENEWAL TERMS.  Always get a right to renew your property at the end of the lease.  You should have a chance to say whether or not you wish to remain in your space but not be obligated to remain in a place that no longer works for you."

Much more from earlier and originals at our LinkedIn Posts

Please email the links to your related articles to - especially the following headline, the subject of our deepening research.  We'll share them as Golden Nuggets for our followers, and notify you when we do, so you can re-share and re-mind your followers of your content. 

LinkedIn • LinkedIn Company Page • Twitter
USA 302KentonJ (GoogleVoice 302-536-8665 to mobile and Hangouts)

  • Quickly consumedmemorable and powerful
  • Linked to the source for much-deeper information
  • Edited and summared showing understanding and thought leadership
  • Mainly shared content from professionals in Commercial Real Estate, Business Consulting, and Thought and Executive Leadership
  • Shared on:  LinkedIn (company)LinkedIn (personal)Twitter (automatically from LinkedIn), then highlighted on our Blog (weekly), The News Funnel (weekly), as a LinkedIn article (weekly) and our GoogleGroup InfoPortal (weekly/semi-monthly)
  • Posted originally at
Of those who we've shared, the savvy ones Like or Comment on these "Golden Nuggets" as additional exposure, as well as serve the community and remind their followers of their wisdom.