Gotta get used to writing dates

7:00 – 5:00 ASRC PhD NASA

• Continuing my deep neural network writeup
• Continuing Holt-Winters work with Aaron
• Continuing to read Clockwork Muse
  • Martindale spends the book talking about poetry, but I’m listening to Kind of Blue right now and I realize that Jazz is similar. The thought leaders are in some state where they are paying attention to each other and not much else. That’s how we get a trajectory that leads to Bitches Brew.
  • I think this is probably a generally applicable pattern. The thing that I need to think through is how a small group of highly creative people in what could be described as an echo-ish chamber differs from mass activity in a large attractor like authoritarianism.
  • We want to know how many dimensions are needed to account for the similarities among poets. Fortunately, once we have correlated all of the poets with one another, a procedure called multidimensional scaling will tell us just this.* Multidimensional scaling tells us that the twenty-one French poets differ along three· main dimensions. These three dimensions account for 94 percent of the similarity matrix. (Page 114)
• I have spent the day doing PHENOMENALLY STUPID MATH (Holt exponential smoothing with drecksnest damping). Excel file: ExponentialSmoothing2

7:00 – 4:30 ASRC NASA

• Set up appt for physical
• This is fabulous! Seeing Theory
  • Seeing Theory was created by Daniel Kunin while an undergraduate at Brown University. The goal of this website is to make statistics more accessible through interactive visualizations (designed using Mike Bostock’s JavaScript library D3.js).
• Working on cleaning up and validating my Very Simple Perceptron class. I think I’m going to write the whole thing up as its own blog post
• Why Trump Reigns as King Cyrus
  • This isn’t the religious right we thought we knew. The Christian nationalist movement today is authoritarian, paranoid and patriarchal at its core. They aren’t fighting a culture war. They’re making a direct attack on democracy itself.

7:00 – 4:30 ASRC PhD/NASA/NOAA

• Spatial Representations in the Human Brain
  • While extensive research on the neurophysiology of spatial memory has been carried out in rodents, memory research in humans had traditionally focused on more abstract, language-based tasks. Recent studies have begun to address this gap using virtual navigation tasks in combination with electrophysiological recordings in humans. These studies suggest that the human medial temporal lobe (MTL) is equipped with a population of place and grid cells similar to that previously observed in the rodent brain. Furthermore, theta oscillations have been linked to spatial navigation and, more specifically, to the encoding and retrieval of spatial information. While some studies suggest a single navigational theta rhythm which is of lower frequency in humans than rodents, other studies advocate for the existence of two functionally distinct delta–theta frequency bands involved in both spatial and episodic memory. Despite the general consensus between rodent and human electrophysiology, behavioral work in humans does not unequivocally support the use of a metric Euclidean map for navigation. Formal models of navigational behavior, which specifically consider the spatial scale of the environment and complementary learning mechanisms, may help to better understand different navigational strategies and their neurophysiological mechanisms. Finally, the functional overlap of spatial and declarative memory in the MTL calls for a unified theory of MTL function. Such a theory will critically rely upon linking task-related phenomena at multiple temporal and spatial scales. Understanding how single cell responses relate to ongoing theta oscillations during both the encoding and retrieval of spatial and non-spatial associations appears to be key toward developing a more mechanistic understanding of memory processes in the MTL.
• Three Kinds of Spatial Cognition
  • Nora S. Newcombe (Scholar)
  • Spatial cognition is often (but wrongly) conceptualized as a single domain of cognition. However, humans function in more than one way in the spatial world. We navigate, as do all mobile animals, but we also manipulate objects using distinctive hands with opposable thumbs, unlike other species. In fact, an important characteristic of human adaptation is the ability to invent tools. Of course, another central asset is human symbolic ability, which includes the ability to spatialize thought in abstractions such as maps, graphs, and analogies. Thus, there are at least three kinds of spatial cognition with three separable functions. Navigation involves moving around the environment to find food and shelter, and to avoid danger. It draws on several interconnected neural subsystems that track movement and encode the location of external entities with respect to each other and the moving self (i.e., extrinsic coding), and it integrates these inputs to achieve best‐possible estimates. Human navigation is characterized by a great deal of individual variation. Tool use and invention involves the mental representation and transformation of the shapes of objects (i.e., intrinsic coding). It relies on substantially different neural subsystems than navigation. Like navigation, it shows marked individual differences, which are related to variations in learning in science, technology, engineering, and mathematics (STEM). Spatialization is an aspect of human symbolic skill that cuts across multiple cognitive domains and involves many kinds of spatial symbol systems, including language, metaphor, analogy, gesture, sketches, diagrams, graphs, maps, and mental images. These spatial symbol systems are vital to many kinds of learning, including in STEM. Future research on human spatial cognition needs to further delineate the origins, development, neural substrates, variability, and malleability of navigation, tool use, and abstract spatial thinking, as well as their interconnections to each other and to other cognitive skills.
• A little bit adding to Normal Accidents notes
• Working on saving out to history and item table
• Turns out that if you want to retrieve floats from a postgres table using psycopg2, you have to register a custom handler:

  DEC2FLOAT = psycopg2.extensions.new_type(
      psycopg2.extensions.DECIMAL.values,
      'DEC2FLOAT',
      lambda value, curs: float(value) if value is not None else None)
  psycopg2.extensions.register_type(DEC2FLOAT)

• Learned about dollar-quoting as per https://www.postgresql.org/docs/current/sql-syntax-lexical.html#SQL-SYNTAX-DOLLAR-QUOTING
• To execute big inserts with psycopg2, you need to set autocommit = True

  self.conn = psycopg2.connect(config_str)
  self.conn.autocommit = True

• And you should use try/catch

  def query_no_result(self, sql: str) -> bool:
      try:
          self.cursor.execute(sql)
          desc = self.cursor.statusmessage
          return True
      except:
          print("unable to execute: {}".format(sql))
          return False

7:00 – 4:30 ASRC PhD/NASA

• I think the IEEE paper with Antonio should be something on the math behind diversity/chaos <-> ensembles <-> hierarchies/stampedes
• Continuing with Normal Accidents
• Thinking about costly signalling (economics, and more generally) WRT stampede theory. It’s a form of friction. Proof-of-work could be adaptively added to a system to inhibit stampedes?
  • In contract theory, signalling is the idea that one party (termed the agent) credibly conveys some information about itself to another party (the principal). For example, in Michael Spence’s job-market signalling model, (potential) employees send a signal about their ability level to the employer by acquiring education credentials. The informational value of the credential comes from the fact that the employer believes the credential is positively correlated with having greater ability and difficult for low ability employees to obtain. Thus the credential enables the employer to reliably distinguish low ability workers from high ability workers.
• Tweak optimizer. Change callback method to replace_with_your_method(), and add some documentation on how to use the classes
• Move clustering paper to LaTex folder – done!
• Pinged Antonio about the new paper ideas
• More predictive analytics? Kind of – setting up to read and write out time series to db
• Upgrading my postgreSQL, to read/write from a table
  • Installed drivers and python packages
  • Reading and writing in the IDE
    • Created and populated a dummy data table
  • Reading in python. Writing is next
• I was thinking that if we run out of D&D data for other map configurations, that I can use the current data with some repurposing (replace orc with xxx) to create conversations about other environments using human speech. Kind of like PCA for DNA.

7:00 – 4:30 ASRC PhD/NASA

• Sent Greg a couple of quick notes on using CNNs to match bacteria to phages.
• Continuing with Normal Accidents
• A Digital Test of the News: Checking the Web for Public Facts – Workshop report, December 2018
  • The Digital Test of the News workshop brought together digital sociologists, data visualisation and new media researchers at the Centre for Interdisciplinary Methodologies at the University of Warwick on 8 and 9 May 2018. The workshop is part of a broader research collaboration between the Centre for Interdisciplinary Methodologies and the Public Data Lab which investigates the changing nature of public knowledge formation in digital societies and develops inventive methods to capture and visualise knowledge dynamics online. Below we outline the workshop’s aims and outcomes.
• Added plots to the NN code. Everything seems to look right. Looking at the individual weights in a layer is very informative. Need to add this kind of plotting to our keras code somehow:

  

  

  

  
  
• Changing the coherence code so that the row values are zero or one. Actually, as the amount of data grows, BOW is getting more useful raw. This spreadsheet shows all posts, including the DM. Note that the word frequency is power law (R squared = .9352):
• Started the optimizer and excel utils classes
• NOAA meeting
• NOAA meeting 2
• Some thoughts from Aaron on our initial ML approach
  • I think for the first pass we do 1-2 models based on contract type focused on the $500k+ award contracts (120ish total).
  • We construct the inputs like sentences in the word generation LSTM with padded lengths equal to the longest running contract of that type, sorted by length. The model can be tested against current contracts held out in a test set using point by point prediction so we can show accuracy of the model against existing data and use that to set our accuracy threshold.
  • My guess is this will be at least an ORF/PAC model (two different primary contract types) which we can work on tuning with the learning_optimizer.py to get as accurate as possible in the timeframe we have.
  • One of the things we advertise as “next steps” is a detailed analysis of contracts based on similarity measures to identify a series of more accurate models. We can pair this with additional models such as Exponential Smoothing and ARIMA which use fundamentally the exact same pipeline.
  • The GUI will be plumbed up to show these analytic outputs on a per contract basis and we can show by the end of January a simple linear model and the LSTM model to demonstrate how Exponential Smoothing / ARIMA or model averages could be displayed. Once we have these outputs we can take the top 5 highest predicted UDO and display in a summary page so they can use those as a launching off point.
  • If we do it this way it means we only have to focus on the completion of TimeSeriesML LSTM and it’s data pipeline with a maximum of 2 initial models (contract type). I think that is a far more reasonable thing to complete in the timeframe and should still be really exciting to show off.