The role of Statistics in the world of Big Data

class: center, middle, inverse, title-slide

# The role of Statistics in the world of Big Data
### Salvador, 2018

---

class: class: center, middle

# Data janitor

---
class: center, middle

---
class: middle, center

### Food for thought

---

## New era?

- Data variety increased.
- Important data today:

---

## Deep Learning

- Recent popularity and hype.
- Astonishing case studies.
- Different terminology

---

## Problems

- Many (too many?) people using it acritically.

- Market is demanding.

- We don't learn that in undergrad courses (or do we?)

### Are statisticians becoming obsolete?

---
class: inverse, middle, center

# Depression

---

## Marketing and venn diagrams

---

## Marketing and venn diagrams

---

## Marketing and venn diagrams

---

## Depression

---

## Depression 2

---
class:inverse

## Don't panic!

---

## Don't panic!

- The definition of Data Science does not really matter.

- What we should discuss is how to actually **DO** data science

---

## Don't panic!

### There are many false cognates.

### Our course IS useful.

### We need to adapt our mindset and develop ourselves.

---

## Logistic regression

`$$\mathbb E(Y_i) = g^{-1}(\alpha + x_i\beta)$$`

---

## Deep learning

`$$f(x) = \sigma(wx + b)$$`

- Coincidence?

---
class: inverse, center, middle

# What's my role?

---
class: center, middle

# we give data science direction

In a Gradient Descent optimization, the **gradient** gives the direction, and the **learning rate** gives the size of the step.

`$$\beta_{\text{new}} = \beta_{\text{old}} - \alpha\nabla_\beta(\text{loss})$$`

In data science, statisticians' role is to be the **gradient**, and computer scientists' role is to be the **learning rate**.

`$$DS_{\text{new}} = DS_{\text{old}} - cs * stat (errors)$$`

---

## Where we are

---

## Including data science

---
class: center, middle

# What we want?

---

## More registered statisticians (CONRE)

---

## More undergrad courses, less evasion

---

## Work with other data scientists

---
class: inverse, center, middle

# What should we do?

---

# What do you want?

## LEARN (L)

- Learn many things, write papers, exercise your curiosity

## RESOLVE (R)

- Earn money, create your startup, raise a family

## SHARE (S)

- Share your profession, be relevant on the web, help the community

---

# Focus

--
## 1. [LSR] Integrate communities

--
## 2. [SLR] Be relevant on the web

--
## 3. [LRS] Study, learn, update, use R (and python)

--
## 4. [RSL] Use what the university give to you

--
## 5. [RLS] Be relevant in the university

---

# Stalk me

- CONRE-3: [jtrecenti@conre3.org.br](mailto:jtrecenti@conre3.org.br)

Pages:

- https://curso-r.com
- https://abj.org.br

Slides: https://jtrecenti.github.io/slides/ufba-rt/