"To develop a complete mind: Study the science of art; Study the art of science. Learn how to see.
Realize that everything connects to everything else."

Experience & Skills

  • Machine Learning,
  • Web Development,
  • Physics and Astrophysics,
  • Drawing, Painting, and Design
  • Data Science,
  • Teaching and Public Speaking
about me

Filis
Coba

Data Scientist/Physicist

I have a master's degree in physics from Old Dominion University under the advisment of Declan de Paor and researched the interior of Mars and how disruptive technology affects undergraduate astronomy learning.

I participated in a 15 week full time, immersive data science bootcamp with Flatiron School. Before that, I participated in a 6-month data analytics and visualization bootcamp with George Washington University and have shifted my focus to machine learning and data science. I believe these are our future and given my skills in science, art, and teaching I feel I can contribute to this endevour.

When I am not doing science or art, I enjoy hiking and Kendo martial arts.

Filis Coba Washington, DC

Classifying Countries Based on Life Expectancy

In this project, we process data from the World Health Organization on a country-by-country basis and develop a model to predict whether a country's life expectancy is greater than or less than the world's median life expectancy.

Project Details

  • Flatiron School
  • Tech:Pandas, Altair, scikit-learn

Details:

  1. Clean the data of NaN's and check that the data for each country is stationary.
  2. Altair visual required getting country code data 
    source = alt.topo_feature(data.world_110m.url,'countries')

map_plot = alt.Chart(source).mark_geoshape().encode(
  color=alt.Color('Life expectancy :Q', legend=alt.Legend(title='Years')),
  tooltip='tooltip:N'
).transform_lookup(
 lookup='id',
 from_=alt.LookupData(df_2015_map, 'id', ['Life expectancy ', 'tooltip'])
).project(
 type='equirectangular'
).properties(
  width=900,
  height=540,
  title=('Life Expectency in Years')
)
  3. Implement (a) decision tree and (b) random forest classifier (91% and 93% accuracy, respectively).
  4. Determine from the models that schooling is the most important feature, followed by HIV/AIDS prevalence. Our conclusions are careful not to confuse these features as causal.
Visit GitHub
Prev
Next

Let's Explore !


Filis Coba Washington, DC