🧪 Synthetische Daten (in der Verwaltung 🗃️)
+🛠️ Arbeiitsversion - Work in progress !
Ein Wiki zu synthetischen Daten und Differential Privacy und deren potenziellen Nutzen in der Verwaltung.
Dieses Wiki zielt auf eine praxisnahe und anwendungsorientierte Einführung in das Thema synthetische Daten ab. Es dient hoffentlich allen, die..:
diff --git a/nutzen.html b/nutzen.html
index 6fd1343..022d5e3 100644
--- a/nutzen.html
+++ b/nutzen.html
@@ -35,6 +35,7 @@
+
diff --git "a/qualit\303\244tsbewertung.html" "b/qualit\303\244tsbewertung.html"
index 913675a..c082120 100644
--- "a/qualit\303\244tsbewertung.html"
+++ "b/qualit\303\244tsbewertung.html"
@@ -35,6 +35,7 @@
+
@@ -66,7 +67,7 @@
-
+
@@ -153,14 +154,14 @@
-- Synthetische Daten in der Verwaltung
+- Synthetische Daten
- Besonderheiten
-- Praxisbeispiele
+- Praxisbeispiele in der Verwaltung
- Qualitätsbewertung
- Datenschutz
- Datenauswertung
@@ -408,7 +409,7 @@ Qualitätsbewertung
+
diff --git a/searchindex.js b/searchindex.js
index 3c3d291..4a5743c 100644
--- a/searchindex.js
+++ b/searchindex.js
@@ -1 +1 @@
-Search.setIndex({"docnames": ["acknowledgements", "besonderheiten", "datenauswertung", "datenschutz", "diff_privacy", "diff_privacy_sd", "diff_privacy_sd_gan", "dummy_data", "einleitung", "intro", "nutzen", "qualit\u00e4tsbewertung"], "filenames": ["acknowledgements.md", "besonderheiten.ipynb", "datenauswertung.ipynb", "datenschutz.ipynb", "diff_privacy.ipynb", "diff_privacy_sd.ipynb", "diff_privacy_sd_gan.ipynb", "dummy_data.ipynb", "einleitung.ipynb", "intro.md", "nutzen.md", "qualit\u00e4tsbewertung.ipynb"], "titles": ["Acknowledgements", "Besonderheiten", "Datenauswertung", "Datenschutz", "Differential Privacy", "Beispiel Differential Privacy", "Synthetische Daten mit KI", "Generierung von Dummy Daten", "Synthetische Daten", "\ud83e\uddea Synthetische Daten (in der Verwaltung \ud83d\uddc3\ufe0f)", "Praxisbeispiele in der Verwaltung", "Qualit\u00e4tsbewertung"], "terms": {"mani": 0, "brilliant": 0, "peopl": 0, "idea": 0, "enabl": 0, "thi": [0, 7], "collect": 0, "i": [0, 5], "am": [0, 10], "veri": 0, "great": 0, "support": [0, 7], "advic": 0, "thank": 0, "all": [0, 4], "author": 0, "sourc": 0, "were": 0, "us": 0, "book": [0, 5], "wa": [0, 8], "inspir": 0, "tutori": 0, "paper": 0, "other": 0, "jupyt": 0, "check": [0, 7], "them": 0, "out": 0, "patrick": 0, "mineault": 0, "The": [0, 7], "good": 0, "research": [0, 10], "code": 0, "handbook": 0, "intuit": 0, "joseph": [0, 5], "p": [0, 5], "Near": [0, 5], "chik\u00e9": [0, 5], "abuah": [0, 5], "program": 0, "differenti": [0, 6, 9], "privaci": [0, 6, 9], "easi": 0, "follow": 0, "explan": 0, "approach": 0, "teach": 0, "section": 0, "build": 0, "mainli": 0, "exampl": 0, "tba": 0, "cite": 0, "work": [0, 7], "includ": [0, 7], "hernandez": [0, 11], "et": [0, 11], "al": [0, 2, 4, 5, 7, 8, 10, 11], "2021": [0, 11], "standardis": [0, 11], "metric": [0, 11], "method": [0, 11], "synthet": [0, 6, 7, 11], "tabular": [0, 11], "data": [0, 2, 5, 6, 7, 8, 10, 11], "evalu": [0, 11], "techrxiv": [0, 11], "gener": [0, 7, 8], "jupyterbook": 0, "which": 0, "sphinx": 0, "synthetisch": [1, 2, 10], "unterscheiden": [1, 4], "sich": [1, 3, 4, 6], "nach": [1, 5, 6], "dem": [1, 10], "grad": 1, "der": [1, 2, 3, 4, 5, 6, 7, 8, 11], "wiedergabetreu": [1, 11], "sehen": [1, 9], "au": [1, 3, 5, 8, 10], "und": [1, 2, 3, 4, 6, 7, 8, 9, 11], "f\u00fchlen": 1, "an": [1, 2, 4, 5, 6], "wie": [1, 2, 3, 8, 10], "echt": [1, 8], "2": [1, 5, 6, 7, 8], "enthalten": 1, "kein": [1, 4, 7], "personenbezogenen": 1, "informationen": [1, 2, 3], "da": [1, 2, 3, 4, 5, 6, 7, 9, 10], "sie": [1, 4], "zuf\u00e4lligen": [1, 8], "zahlen": [1, 8], "generiert": [1, 8], "werden": [1, 3, 7, 8, 11], "geben": 2, "prei": 2, "oft": 2, "form": 2, "statistiken": 2, "aggregiert": 2, "repr\u00e4sentationen": 2, "nutzen": [2, 7, 9], "indem": 2, "informationsabfragen": 2, "ein": [1, 2, 3, 4, 5, 6, 9, 11], "datengrundlag": [2, 3, 8, 10], "stellen": [2, 5], "viel": 2, "kinder": [2, 5], "sind": [2, 4, 5, 6, 8, 9], "im": [2, 4, 6, 7, 8, 10], "zwischen": [2, 5], "9": [2, 5, 8], "jahr": 2, "12": [2, 5, 8], "alt": 2, "anzahl": [2, 5, 7], "pro": [2, 5], "alter": [2, 5], "un": 2, "reichen": 2, "hier": 2, "erstmal": 2, "wert": [2, 7, 8], "bspw": [2, 3, 7, 10], "einem": [2, 3, 4, 5], "balkendiagramm": 2, "nutzung": [2, 6, 10], "einer": [2, 3, 5, 8], "synthetischen": [2, 5, 6, 9, 10, 11], "repr\u00e4sent": [2, 5], "wahrung": 2, "datenschutz": [2, 5, 9, 10], "nachteil": 2, "hilft": 2, "nur": [2, 4], "bei": [2, 6, 7, 8, 10], "dieser": [2, 8], "einen": [2, 4], "spezifischen": [2, 4], "f\u00fcr": [1, 2, 4, 6, 8, 10, 11], "ander": [2, 3, 7], "m\u00fcssen": 2, "neue": 2, "erstellen": [2, 5], "import": [2, 5, 7, 8], "panda": [2, 5, 7, 8], "pd": [2, 5, 7, 8], "matplotlib": [2, 5], "pyplot": [2, 5], "plt": [2, 5], "med_data": [2, 5, 8], "read_excel": [2, 5, 7, 8], "rpad_data_smal": [2, 5, 7, 8], "xlsx": [2, 5, 7, 8], "engin": [2, 5, 8], "openpyxl": [2, 5, 8], "funktion": [2, 4, 5], "zur": [1, 2, 3, 4, 5, 6, 7, 10], "alterspann": [2, 5], "b": [2, 5], "def": [2, 5, 7], "range_queri": [2, 5], "df": [2, 5], "col": [2, 5], "return": [2, 5, 7], "len": [2, 5, 7], "print": [2, 5], "f": [2, 4, 5], "jahren": [2, 5], "ag": [2, 5, 7, 8], "250": [2, 5], "bin": [2, 5], "list": [2, 5, 7], "rang": [2, 5], "0": [2, 5, 7, 8], "20": [2, 5, 7, 8], "count": [2, 5], "1": [2, 5, 7, 8], "xlabel": [2, 5], "ylabel": [2, 5], "bar": [2, 5], "barcontain": [2, 5], "object": [2, 5, 7], "artist": [2, 5], "ergebni": [2, 4, 5, 10], "ist": [2, 4, 7], "genau": 2, "gleich": [2, 4, 5, 8], "egal": 2, "ob": 2, "die": [1, 2, 3, 4, 5, 6, 8, 9, 10, 11], "echten": [2, 8, 10], "oder": [2, 4], "privatsph\u00e4r": [3, 4, 6], "folgend": 3, "synonym": 3, "verwendet": 3, "beziehen": 3, "auf": [3, 4, 5, 8, 9, 10], "verhindern": 3, "de": [3, 4, 7], "preisgeben": 3, "von": [3, 4, 5, 6, 11], "sch\u00fctzenswerten": 3, "techniken": 3, "umsetzung": 3, "erlauben": 3, "informieren": 3, "\u00fcber": 3, "trend": 3, "eigenschaften": [3, 8, 9, 10], "ohn": [3, 4, 6], "einzelnen": [3, 4], "personen": 3, "preiszugeben": 3, "beabsichtigen": 3, "ver\u00f6ffentlichung": 3, "daten": [3, 4, 10, 11], "kontrol": 3, "detailgrad": 3, "datensatz": [3, 4, 6, 7, 8], "darunt": 3, "z\u00e4hlen": 3, "nicht": [3, 9], "datensicherheit": 3, "verschl\u00fcsselung": 3, "dies": [3, 9], "verfolgen": 3, "pr\u00e4vention": 3, "den": [3, 4, 6, 10], "zugriff": 3, "erlaubt": 4, "steuerung": 4, "grade": 4, "gold": 4, "standard": [4, 6], "beim": 4, "schutz": 4, "vor": 4, "reidentifik": 4, "angriffen": 4, "umsetzt": 4, "gilt": 4, "konform": [4, 9], "wenn": 4, "nachbar": 4, "datens\u00e4tz": [4, 10], "x": [4, 7], "m\u00f6glichen": 4, "output": 4, "s": 4, "frac": 4, "mathsf": 4, "pr": 4, "leq": 4, "e": [4, 7], "epsilon": [4, 5], "wird": [4, 7], "bleiben": 4, "mit": [4, 7, 9, 11], "individuum": 4, "dabei": 4, "zwei": 4, "benachbart": 4, "um": [4, 11], "eintrag": 4, "nennt": 4, "paramet": 4, "bzw": 4, "budget": 4, "erlaub": 4, "meng": 4, "zu": [4, 5, 6, 7, 9, 10], "steuern": 4, "klein": [4, 10], "setzt": 4, "hohe": [4, 11], "level": 4, "gro\u00dfe": 4, "erfordert": [4, 11], "wenig": [4, 7], "\u00e4hnlich": 4, "bieten": 4, "damit": [4, 8], "differenziel": 4, "privat": 4, "algorithmu": [4, 5], "abfrag": [4, 9], "verh\u00e4lt": 4, "ungef\u00e4hr": 4, "\u00e4hnlichen": 4, "datens\u00e4tzen": 4, "di": 4, "versichert": 4, "dass": [4, 9], "pr\u00e4senz": 4, "abwesenheit": 4, "keinen": 4, "signifikanten": 4, "einfluss": 4, "hat": 4, "es": [4, 6, 9], "r\u00fcckschl\u00fcsse": 4, "merkmal": 4, "einzeln": 4, "individuen": 4, "m\u00f6glich": 4, "besonderheit": 4, "bezug": 4, "unterschi": 4, "k": 4, "anonym": 4, "erstellt": [4, 5], "wir": [5, 6], "beginnen": 5, "wieder": 5, "unser": 5, "kidner": 5, "numpi": [5, 7], "np": [5, 7], "dazu": 5, "ergibt": [5, 6], "range_query_synth": 5, "syn_rep": 5, "total": 5, "hinzuf\u00fcgen": 5, "rauschen": 5, "jeder": 5, "beobachtung": 5, "laplace_mech": 5, "v": 5, "sensit": 5, "random": [5, 7], "laplac": 5, "loc": 5, "scale": 5, "dp_syn_rep": 5, "c": 5, "256": 5, "07787330314517": 5, "zun\u00e4chst": 5, "simpl": 5, "altersgrupp": 5, "normalisieren": 5, "wahrscheinlichkeit": 5, "interpretieren": 5, "dp_syn_rep_nn": 5, "clip": 5, "none": [5, 7], "syn_norm": 5, "sum": 5, "9999999999999998": 5, "zuf\u00e4llig": 5, "stichproben": 5, "altersangaben": 5, "ziehen": 5, "gewichtet": 5, "berechneten": 5, "stichprob": 5, "5": [5, 7, 8], "werten": [5, 8], "\u00fcberpr\u00fcfung": 5, "gen_sampl": 5, "n": 5, "choic": 5, "syn_data": 5, "datafram": 5, "column": [5, 7], "14": [5, 8], "3": [5, 6, 7, 8], "4": [5, 6, 8], "8": [5, 7], "1000": 5, "visualisierung": 5, "verteilung": [5, 8], "diagramm": 5, "10000": 5, "hist": 5, "rwidth": 5, "90": 5, "arrai": 5, "46": 5, "25": [5, 8], "48": 5, "153": 5, "266": 5, "311": 5, "247": 5, "517": 5, "586": 5, "893": 5, "1140": 5, "1286": 5, "1207": 5, "759": 5, "855": 5, "838": 5, "466": 5, "341": 5, "16": [5, 8], "6": [5, 8], "7": [5, 7, 8], "10": [5, 7, 8], "11": [5, 8], "13": 5, "15": [5, 8], "17": 5, "18": [5, 8], "19": 5, "konformen": 5, "haben": [5, 8, 9, 10], "halten": 5, "auch": 5, "festgesetzt": 5, "ma\u00df": 5, "quell": [1, 5], "vorlag": 5, "A": 5, "about": 5, "programm": 5, "By": 5, "gefahren": 6, "teilbar": 6, "aktuel": 6, "dsgvo": 6, "eigentlich": 6, "ja": 6, "aber": [6, 8], "besteht": 6, "risiko": 6, "durch": 6, "fehler": 6, "generierungsprozess": 6, "gegen": 6, "versto\u00dfen": 6, "dadurch": 6, "bedarf": 6, "regulierung": 6, "einheitlichen": 6, "arbeitsgrupp": 6, "industri": 6, "connect": 6, "ieee": 6, "associ": 6, "erarbeitet": 6, "vorschl\u00e4gen": 6, "privatsph\u00e4restandard": 6, "qualit\u00e4tsstandard": 6, "generierten": [6, 8], "risikofrei": 6, "nein": 6, "k\u00f6nnen": [6, 8, 10], "integrieren": 6, "eingehalten": 6, "kontrolliert": 6, "challeng": 6, "nation": 6, "institut": 6, "technolog": 6, "dp": 6, "algorithmen": 6, "generierung": [1, 6, 10, 11], "gan": 6, "version": 6, "belegt": 6, "personenbezug": 7, "format": 7, "struktur": [7, 8], "bleibt": 7, "erhalten": 7, "ziel": [7, 10], "teilen": 7, "verprobung": [7, 10], "forschungside": 7, "softwar": [7, 10], "etc": 7, "load": 7, "necessari": 7, "packag": 7, "datetim": 7, "sy": 7, "from": [7, 8], "date": 7, "disclaim": 7, "datenformat": 7, "einbauen": 7, "note": 7, "apart": 7, "microsoft": 7, "excel": 7, "onli": 7, "two": 7, "dimension": 7, "current": 7, "automat": 7, "upload": 7, "each": 7, "ha": 7, "separ": 7, "variabl": 7, "oppos": 7, "being": 7, "contain": 7, "row": 7, "read_data": 7, "endswith": 7, "csv": 7, "txt": 7, "read_csv": 7, "elif": 7, "xl": 7, "dictionari": 7, "sheet_nam": 7, "kei": 7, "name_of_sheet": 7, "simple_data_fram": 7, "els": 7, "sas7bdat": 7, "read_sa": 7, "sav": 7, "read_spss": 7, "dta": 7, "read_stata": 7, "pkl": 7, "read_pickl": 7, "rais": 7, "except": 7, "sorri": 7, "type": 7, "try": 7, "convert": 7, "numer": 7, "categor": 7, "na": 7, "string": 7, "pattern": 7, "abbilden": 7, "file_path": 7, "original_data": 7, "dtype": 7, "id": [7, 8], "int64": 7, "float64": 7, "bmi": [7, 8], "sex": [7, 8], "height": [7, 8], "weight": [7, 8], "length_of_stai": [7, 8], "manag": [7, 8], "sever": [7, 8], "alvarado_scor": [7, 8], "paedriatic_appendicitis_scor": [7, 8], "dfn": 7, "convert_dtyp": 7, "python": 7, "sheet": 7, "einf\u00fcgen": 7, "kategorisch": 7, "variablen": 7, "sollten": 7, "per": [7, 8], "hand": 7, "definiert": 7, "wissen": 7, "welcher": 7, "datentyp": 7, "unterschiedlich": 7, "insgesamt": 7, "indik": 7, "kleiner": 7, "gesch\u00e4tzt": 7, "beobachtungen": 7, "kategori": 7, "erkannt": 7, "anpassen": 7, "schwell": 7, "categori": 7, "name": [7, 8], "item": 7, "unique_count": 7, "uniqu": 7, "shape": [7, 8], "total_count": 7, "astyp": 7, "problem": 7, "lenght_of_stai": 7, "valu": 7, "780": 7, "threshold": 7, "m\u00fcsste": 7, "02": 7, "liegen": 7, "diesem": 7, "fall": 7, "fine": 7, "evt": 7, "gestaffelt": 7, "ab": [7, 9], "100": 7, "zahl": 7, "udn": 7, "50": 7, "In": 7, "final": 7, "step": 7, "we": 7, "set": 7, "have": 7, "same": 7, "structur": 7, "number": 7, "your": 7, "origin": 7, "freuenci": 7, "preserv": 7, "albeit": 7, "introduct": 7, "some": 7, "nois": 7, "found": 7, "larg": 7, "similar": 7, "g": 7, "frequenc": 7, "new": 7, "append": 7, "its": 7, "save": 7, "directori": 7, "select_dtyp": 7, "new_col": 7, "seri": 7, "replac": 7, "nan": [7, 8], "nameerror": 7, "traceback": 7, "most": 7, "recent": 7, "call": 7, "last": 7, "cell": 7, "line": 7, "defin": 7, "creat": 7, "synth": 7, "between": 7, "real": [7, 8], "remain": 7, "var": 7, "length": 7, "detect": 7, "recreat": 7, "adress": 7, "faker": 7, "adresslin": 7, "k\u00fcnstlich": [8, 10], "erzeugt": 8, "ihrer": 8, "gesamtheit": 8, "statistischen": [8, 10], "datenstichprob": 8, "gemessenen": 8, "nachbildet": 8, "modellierung": 8, "approxim": 8, "erfordern": 8, "model": 8, "produziert": 8, "aussehen": 8, "bedeutung": 8, "tragen": 8, "datenoperationen": 8, "ergebniss": 8, "liefern": 8, "keiner": 8, "realen": 8, "person": 8, "zugeordnet": 8, "perfekt": 8, "proxi": 8, "urspr\u00fcnglich": 8, "kann": [8, 11], "statt": 8, "eingesetzt": 8, "sdv": 8, "metadata": 8, "singletablemetadata": 8, "lite": 8, "singletablepreset": 8, "head": 8, "7898": 8, "68": 8, "femal": 8, "148": 8, "37": 8, "conserv": 8, "uncompl": 8, "6952": 8, "31": 8, "male": 8, "147": 8, "69": 8, "7909": 8, "23": 8, "163": 8, "62": 8, "8163": 8, "165": 8, "56": 8, "1933": 8, "08": 8, "45": 8, "080000": 8, "900000": 8, "205": 8, "9035": 8, "350000": 8, "600000": 8, "240": 8, "9700": 8, "267625": 8, "335466": 8, "detect_from_datafram": 8, "synthes": 8, "fast_ml": 8, "fit": 8, "med_data_synth": 8, "sampl": 8, "num_row": 8, "7109": 8, "236162": 8, "21": 8, "136330": 8, "176": 8, "644801": 8, "63": 8, "284976": 8, "942523": 8, "087310": 8, "049639": 8, "5765": 8, "050711": 8, "566346": 8, "103": 8, "906050": 8, "438980": 8, "257973": 8, "905003": 8, "881729": 8, "2402": 8, "727097": 8, "196794": 8, "138": 8, "056399": 8, "037305": 8, "902269": 8, "000000": 8, "364234": 8, "8480": 8, "964649": 8, "336835": 8, "143": 8, "513145": 8, "516364": 8, "043838": 8, "704921": 8, "327257": 8, "3978": 8, "201880": 8, "236390": 8, "144": 8, "011407": 8, "35": 8, "515245": 8, "774677": 8, "770871": 8, "f\u00e4llt": 8, "euch": 8, "spoiler": 8, "numerischen": 8, "zum": [8, 9, 10], "beispiel": [1, 8, 9], "mehr": 8, "nachkommastellen": 8, "originalen": 8, "sg": [], "lunch": [], "talk": [], "teil": [], "wo": [], "einsatz": 10, "sinnvollsten": [], "welch": [], "datenschutzbedenken": [], "m\u00f6glichkeiten": [], "mitigieren": [], "fallen": [], "besonderheiten": 9, "praxisbeispiel": 9, "qualit\u00e4tsbewertung": 9, "datenauswertung": 9, "ki": [9, 10], "acknowledg": 9, "praxi": 10, "dummi": 10, "erm\u00f6glichen": 10, "trotz": 10, "datenschutzvorgaben": 10, "versuchsdaten": 10, "softwaretest": 10, "datenbeispiel": 10, "konzepten": 10, "forschungsvorhaben": 10, "produktvorstellungen": 10, "arbeitseb": 10, "projekt": 10, "administr": 10, "uk": 10, "econom": 10, "social": 10, "council": 10, "kritisch": 10, "betrachtung": 10, "datenqualit\u00e4t": 10, "verst\u00e4ndni": 10, "auswirkungen": 10, "vorschl\u00e4g": 10, "ausweitung": 10, "echter": 10, "produkt": 10, "colab": 10, "notebook": 10, "erweitern": 10, "unvollst\u00e4ndig": 10, "erg\u00e4nzen": 10, "bia": 10, "aufheben": 10, "googl": 10, "cloud": 10, "platform": 10, "gretel": 10, "integr": 10, "dienstleistung": 10, "mlop": 10, "plattform": 10, "vertex": 10, "ai": 10, "verbesserung": 10, "train": 10, "ml": 10, "modellen": 10, "sensibl": 10, "internen": 10, "verarbeitung": 10, "scienc": 10, "team": 10, "forschungszwecken": 10, "bereitstellen": 10, "intelligenz": 10, "forschungsdatenzentrum": 10, "bfarm": 10, "erforschung": 10, "anonymisierungsm\u00f6glichkeiten": 10, "readi": 10, "fdz": 10, "vorhaben": 10, "gesundheitsdaten": 10, "erzeugen": 10, "jedoch": 10, "gleichen": 10, "projektbeschreibung": 10, "insbesonder": 11, "forschungsdaten": [], "qualit\u00e4t": 11, "u": 11, "folgenden": 11, "metriken": 11, "gemessen": 11, "\u00fcbersicht": 11, "stammt": 11, "wurd": 11, "eigen": 11, "angaben": 11, "erg\u00e4nzt": 11, "qualit\u00e4tsmetriken": 11, "baum": 11, "englisch": 11, "forschung": 11, "architektur": 1, "neuronalen": 1, "netz": 1, "zufallszahlen": 1, "http": 1, "datacebo": 1, "com": 1, "blog": 1, "interpret": 1, "ctgan": 1, "progress": 1, "wiki": 9, "deren": 9, "potenziellen": 9, "zielt": 9, "praxisnah": 9, "anwendungsorientiert": 9, "einf\u00fchrung": 9, "thema": 9, "dient": 9, "hoffentlich": 9, "allen": 9, "neugierig": 9, "tats\u00e4chlich": 9, "anwendungen": 9, "wollen": 9, "ganz": 9, "\u00fcberzeugt": 9, "sinnvol": 9, "skeptisch": 9, "statistisch": 9, "nachbildbar": 9, "zweifel": 9, "anleitung": 9, "selbermachen": 9, "suchen": 9}, "objects": {}, "objtypes": {}, "objnames": {}, "titleterms": {"acknowledg": 0, "refer": 0, "besonderheiten": 1, "welch": 1, "arten": 1, "von": [1, 2, 7, 10], "synthetischen": 1, "daten": [1, 2, 5, 6, 7, 8, 9], "gibt": 1, "es": 1, "wa": 1, "macht": 1, "den": 1, "einsatz": 1, "spannend": 1, "datenauswertung": 2, "warum": 2, "brauchen": 2, "wir": 2, "beispiel": [2, 5], "abfrag": [2, 5], "h\u00e4ufigkeiten": 2, "hinwei": 2, "datenschutz": 3, "begriff": 3, "ziel": 3, "differenti": [4, 5], "privaci": [4, 5], "mathematisch": 4, "definit": [4, 8], "intuit": 4, "mit": [5, 6], "und": [5, 10], "synthetisch": [5, 6, 8, 9], "datenschutzkonform": 5, "ki": 6, "bestehend": 6, "regulierungen": 6, "forschungsergebniss": 6, "platz": 6, "1": 6, "5": 6, "generierung": 7, "dummi": 7, "write": 7, "test": 7, "function": 7, "see": 7, "sa": 7, "stata": 7, "file": 7, "run": 7, "properli": 7, "der": [9, 10], "verwaltung": [9, 10], "ein": 8, "datenbeispiel": 8, "teaser": [], "diskuss": [], "praxisbeispiel": 10, "struktur": 10, "aufbau": 10, "datens\u00e4tzen": 10, "veranschaulichen": 10, "echt": 10, "datengrundlagen": 10, "aufwerten": 10, "sch\u00fctzenswert": 10, "teilen": 10, "qualit\u00e4tsbewertung": 11}, "envversion": {"sphinx.domains.c": 2, "sphinx.domains.changeset": 1, "sphinx.domains.citation": 1, "sphinx.domains.cpp": 6, "sphinx.domains.index": 1, "sphinx.domains.javascript": 2, "sphinx.domains.math": 2, "sphinx.domains.python": 3, "sphinx.domains.rst": 2, "sphinx.domains.std": 2, "sphinx.ext.intersphinx": 1, "sphinxcontrib.bibtex": 9, "sphinx": 56}})
\ No newline at end of file
+Search.setIndex({"docnames": ["acknowledgements", "besonderheiten", "datenauswertung", "datenschutz", "diff_privacy", "diff_privacy_sd", "diff_privacy_sd_gan", "dummy_data", "einleitung", "intro", "nutzen", "qualit\u00e4tsbewertung"], "filenames": ["acknowledgements.md", "besonderheiten.ipynb", "datenauswertung.ipynb", "datenschutz.ipynb", "diff_privacy.ipynb", "diff_privacy_sd.ipynb", "diff_privacy_sd_gan.ipynb", "dummy_data.ipynb", "einleitung.ipynb", "intro.md", "nutzen.md", "qualit\u00e4tsbewertung.ipynb"], "titles": ["Acknowledgements", "Besonderheiten", "Datenauswertung", "Datenschutz", "Differential Privacy", "Beispiel Differential Privacy", "Synthetische Daten mit KI", "Generierung von Dummy Daten", "Synthetische Daten", "\ud83e\uddea Synthetische Daten (in der Verwaltung \ud83d\uddc3\ufe0f)", "Praxisbeispiele in der Verwaltung", "Qualit\u00e4tsbewertung"], "terms": {"mani": 0, "brilliant": 0, "peopl": 0, "idea": 0, "enabl": 0, "thi": [0, 7], "collect": 0, "i": [0, 5], "am": [0, 10], "veri": 0, "great": 0, "support": [0, 7], "advic": 0, "thank": 0, "all": [0, 4], "author": 0, "sourc": 0, "were": 0, "us": 0, "book": [0, 5], "wa": [0, 8], "inspir": 0, "tutori": 0, "paper": 0, "other": 0, "jupyt": 0, "check": [0, 7], "them": 0, "out": 0, "patrick": 0, "mineault": 0, "The": [0, 7], "good": 0, "research": [0, 10], "code": 0, "handbook": 0, "intuit": 0, "joseph": [0, 5], "p": [0, 5], "Near": [0, 5], "chik\u00e9": [0, 5], "abuah": [0, 5], "program": 0, "differenti": [0, 6, 9], "privaci": [0, 6, 9], "easi": 0, "follow": 0, "explan": 0, "approach": 0, "teach": 0, "section": 0, "build": 0, "mainli": 0, "exampl": 0, "tba": 0, "cite": 0, "work": [0, 7, 9], "includ": [0, 7], "hernandez": [0, 11], "et": [0, 11], "al": [0, 2, 4, 5, 7, 8, 10, 11], "2021": [0, 11], "standardis": [0, 11], "metric": [0, 11], "method": [0, 11], "synthet": [0, 6, 7, 11], "tabular": [0, 11], "data": [0, 2, 5, 6, 7, 8, 10, 11], "evalu": [0, 11], "techrxiv": [0, 11], "gener": [0, 7, 8], "jupyterbook": 0, "which": 0, "sphinx": 0, "synthetisch": [1, 2, 10], "unterscheiden": [1, 4], "sich": [1, 3, 4, 6], "nach": [1, 5, 6], "dem": [1, 10], "grad": 1, "der": [1, 2, 3, 4, 5, 6, 7, 8, 11], "wiedergabetreu": [1, 11], "sehen": [1, 9], "au": [1, 3, 5, 8, 10], "und": [1, 2, 3, 4, 6, 7, 8, 9, 11], "f\u00fchlen": 1, "an": [1, 2, 4, 5, 6], "wie": [1, 2, 3, 8, 10], "echt": [1, 8], "2": [1, 5, 6, 7, 8], "enthalten": 1, "kein": [1, 4, 7], "personenbezogenen": 1, "informationen": [1, 2, 3], "da": [1, 2, 3, 4, 5, 6, 7, 9, 10], "sie": [1, 4], "zuf\u00e4lligen": [1, 8], "zahlen": [1, 8], "generiert": [1, 8], "werden": [1, 3, 7, 8, 11], "geben": 2, "prei": 2, "oft": 2, "form": 2, "statistiken": 2, "aggregiert": 2, "repr\u00e4sentationen": 2, "nutzen": [2, 7, 9], "indem": 2, "informationsabfragen": 2, "ein": [1, 2, 3, 4, 5, 6, 9, 11], "datengrundlag": [2, 3, 8, 10], "stellen": [2, 5], "viel": 2, "kinder": [2, 5], "sind": [2, 4, 5, 6, 8, 9], "im": [2, 4, 6, 7, 8, 10], "zwischen": [2, 5], "9": [2, 5, 8], "jahr": 2, "12": [2, 5, 8], "alt": 2, "anzahl": [2, 5, 7], "pro": [2, 5], "alter": [2, 5], "un": 2, "reichen": 2, "hier": 2, "erstmal": 2, "wert": [2, 7, 8], "bspw": [2, 3, 7, 10], "einem": [2, 3, 4, 5], "balkendiagramm": 2, "nutzung": [2, 6, 10], "einer": [2, 3, 5, 8], "synthetischen": [2, 5, 6, 9, 10, 11], "repr\u00e4sent": [2, 5], "wahrung": 2, "datenschutz": [2, 5, 9, 10], "nachteil": 2, "hilft": 2, "nur": [2, 4], "bei": [2, 6, 7, 8, 10], "dieser": [2, 8], "einen": [2, 4], "spezifischen": [2, 4], "f\u00fcr": [1, 2, 4, 6, 8, 10, 11], "ander": [2, 3, 7], "m\u00fcssen": 2, "neue": 2, "erstellen": [2, 5], "import": [2, 5, 7, 8], "panda": [2, 5, 7, 8], "pd": [2, 5, 7, 8], "matplotlib": [2, 5], "pyplot": [2, 5], "plt": [2, 5], "med_data": [2, 5, 8], "read_excel": [2, 5, 7, 8], "rpad_data_smal": [2, 5, 7, 8], "xlsx": [2, 5, 7, 8], "engin": [2, 5, 8], "openpyxl": [2, 5, 8], "funktion": [2, 4, 5], "zur": [1, 2, 3, 4, 5, 6, 7, 10], "alterspann": [2, 5], "b": [2, 5], "def": [2, 5, 7], "range_queri": [2, 5], "df": [2, 5], "col": [2, 5], "return": [2, 5, 7], "len": [2, 5, 7], "print": [2, 5], "f": [2, 4, 5], "jahren": [2, 5], "ag": [2, 5, 7, 8], "250": [2, 5], "bin": [2, 5], "list": [2, 5, 7], "rang": [2, 5], "0": [2, 5, 7, 8], "20": [2, 5, 7, 8], "count": [2, 5], "1": [2, 5, 7, 8], "xlabel": [2, 5], "ylabel": [2, 5], "bar": [2, 5], "barcontain": [2, 5], "object": [2, 5, 7], "artist": [2, 5], "ergebni": [2, 4, 5, 10], "ist": [2, 4, 7], "genau": 2, "gleich": [2, 4, 5, 8], "egal": 2, "ob": 2, "die": [1, 2, 3, 4, 5, 6, 8, 9, 10, 11], "echten": [2, 8, 10], "oder": [2, 4], "privatsph\u00e4r": [3, 4, 6], "folgend": 3, "synonym": 3, "verwendet": 3, "beziehen": 3, "auf": [3, 4, 5, 8, 9, 10], "verhindern": 3, "de": [3, 4, 7], "preisgeben": 3, "von": [3, 4, 5, 6, 11], "sch\u00fctzenswerten": 3, "techniken": 3, "umsetzung": 3, "erlauben": 3, "informieren": 3, "\u00fcber": 3, "trend": 3, "eigenschaften": [3, 8, 9, 10], "ohn": [3, 4, 6], "einzelnen": [3, 4], "personen": 3, "preiszugeben": 3, "beabsichtigen": 3, "ver\u00f6ffentlichung": 3, "daten": [3, 4, 10, 11], "kontrol": 3, "detailgrad": 3, "datensatz": [3, 4, 6, 7, 8], "darunt": 3, "z\u00e4hlen": 3, "nicht": [3, 9], "datensicherheit": 3, "verschl\u00fcsselung": 3, "dies": [3, 9], "verfolgen": 3, "pr\u00e4vention": 3, "den": [3, 4, 6, 10], "zugriff": 3, "erlaubt": 4, "steuerung": 4, "grade": 4, "gold": 4, "standard": [4, 6], "beim": 4, "schutz": 4, "vor": 4, "reidentifik": 4, "angriffen": 4, "umsetzt": 4, "gilt": 4, "konform": [4, 9], "wenn": 4, "nachbar": 4, "datens\u00e4tz": [4, 10], "x": [4, 7], "m\u00f6glichen": 4, "output": 4, "s": 4, "frac": 4, "mathsf": 4, "pr": 4, "leq": 4, "e": [4, 7], "epsilon": [4, 5], "wird": [4, 7], "bleiben": 4, "mit": [4, 7, 9, 11], "individuum": 4, "dabei": 4, "zwei": 4, "benachbart": 4, "um": [4, 11], "eintrag": 4, "nennt": 4, "paramet": 4, "bzw": 4, "budget": 4, "erlaub": 4, "meng": 4, "zu": [4, 5, 6, 7, 9, 10], "steuern": 4, "klein": [4, 10], "setzt": 4, "hohe": [4, 11], "level": 4, "gro\u00dfe": 4, "erfordert": [4, 11], "wenig": [4, 7], "\u00e4hnlich": 4, "bieten": 4, "damit": [4, 8], "differenziel": 4, "privat": 4, "algorithmu": [4, 5], "abfrag": [4, 9], "verh\u00e4lt": 4, "ungef\u00e4hr": 4, "\u00e4hnlichen": 4, "datens\u00e4tzen": 4, "di": 4, "versichert": 4, "dass": [4, 9], "pr\u00e4senz": 4, "abwesenheit": 4, "keinen": 4, "signifikanten": 4, "einfluss": 4, "hat": 4, "es": [4, 6, 9], "r\u00fcckschl\u00fcsse": 4, "merkmal": 4, "einzeln": 4, "individuen": 4, "m\u00f6glich": 4, "besonderheit": 4, "bezug": 4, "unterschi": 4, "k": 4, "anonym": 4, "erstellt": [4, 5], "wir": [5, 6], "beginnen": 5, "wieder": 5, "unser": 5, "kidner": 5, "numpi": [5, 7], "np": [5, 7], "dazu": 5, "ergibt": [5, 6], "range_query_synth": 5, "syn_rep": 5, "total": 5, "hinzuf\u00fcgen": 5, "rauschen": 5, "jeder": 5, "beobachtung": 5, "laplace_mech": 5, "v": 5, "sensit": 5, "random": [5, 7], "laplac": 5, "loc": 5, "scale": 5, "dp_syn_rep": 5, "c": 5, "256": 5, "07787330314517": 5, "zun\u00e4chst": 5, "simpl": 5, "altersgrupp": 5, "normalisieren": 5, "wahrscheinlichkeit": 5, "interpretieren": 5, "dp_syn_rep_nn": 5, "clip": 5, "none": [5, 7], "syn_norm": 5, "sum": 5, "9999999999999998": 5, "zuf\u00e4llig": 5, "stichproben": 5, "altersangaben": 5, "ziehen": 5, "gewichtet": 5, "berechneten": 5, "stichprob": 5, "5": [5, 7, 8], "werten": [5, 8], "\u00fcberpr\u00fcfung": 5, "gen_sampl": 5, "n": 5, "choic": 5, "syn_data": 5, "datafram": 5, "column": [5, 7], "14": [5, 8], "3": [5, 6, 7, 8], "4": [5, 6, 8], "8": [5, 7], "1000": 5, "visualisierung": 5, "verteilung": [5, 8], "diagramm": 5, "10000": 5, "hist": 5, "rwidth": 5, "90": 5, "arrai": 5, "46": 5, "25": [5, 8], "48": 5, "153": 5, "266": 5, "311": 5, "247": 5, "517": 5, "586": 5, "893": 5, "1140": 5, "1286": 5, "1207": 5, "759": 5, "855": 5, "838": 5, "466": 5, "341": 5, "16": [5, 8], "6": [5, 8], "7": [5, 7, 8], "10": [5, 7, 8], "11": [5, 8], "13": 5, "15": [5, 8], "17": 5, "18": [5, 8], "19": 5, "konformen": 5, "haben": [5, 8, 9, 10], "halten": 5, "auch": 5, "festgesetzt": 5, "ma\u00df": 5, "quell": [1, 5], "vorlag": 5, "A": 5, "about": 5, "programm": 5, "By": 5, "gefahren": 6, "teilbar": 6, "aktuel": 6, "dsgvo": 6, "eigentlich": 6, "ja": 6, "aber": [6, 8], "besteht": 6, "risiko": 6, "durch": 6, "fehler": 6, "generierungsprozess": 6, "gegen": 6, "versto\u00dfen": 6, "dadurch": 6, "bedarf": 6, "regulierung": 6, "einheitlichen": 6, "arbeitsgrupp": 6, "industri": 6, "connect": 6, "ieee": 6, "associ": 6, "erarbeitet": 6, "vorschl\u00e4gen": 6, "privatsph\u00e4restandard": 6, "qualit\u00e4tsstandard": 6, "generierten": [6, 8], "risikofrei": 6, "nein": 6, "k\u00f6nnen": [6, 8, 10], "integrieren": 6, "eingehalten": 6, "kontrolliert": 6, "challeng": 6, "nation": 6, "institut": 6, "technolog": 6, "dp": 6, "algorithmen": 6, "generierung": [1, 6, 10, 11], "gan": 6, "version": 6, "belegt": 6, "personenbezug": 7, "format": 7, "struktur": [7, 8], "bleibt": 7, "erhalten": 7, "ziel": [7, 10], "teilen": 7, "verprobung": [7, 10], "forschungside": 7, "softwar": [7, 10], "etc": 7, "load": 7, "necessari": 7, "packag": 7, "datetim": 7, "sy": 7, "from": [7, 8], "date": 7, "disclaim": 7, "datenformat": 7, "einbauen": 7, "note": 7, "apart": 7, "microsoft": 7, "excel": 7, "onli": 7, "two": 7, "dimension": 7, "current": 7, "automat": 7, "upload": 7, "each": 7, "ha": 7, "separ": 7, "variabl": 7, "oppos": 7, "being": 7, "contain": 7, "row": 7, "read_data": 7, "endswith": 7, "csv": 7, "txt": 7, "read_csv": 7, "elif": 7, "xl": 7, "dictionari": 7, "sheet_nam": 7, "kei": 7, "name_of_sheet": 7, "simple_data_fram": 7, "els": 7, "sas7bdat": 7, "read_sa": 7, "sav": 7, "read_spss": 7, "dta": 7, "read_stata": 7, "pkl": 7, "read_pickl": 7, "rais": 7, "except": 7, "sorri": 7, "type": 7, "try": 7, "convert": 7, "numer": 7, "categor": 7, "na": 7, "string": 7, "pattern": 7, "abbilden": 7, "file_path": 7, "original_data": 7, "dtype": 7, "id": [7, 8], "int64": 7, "float64": 7, "bmi": [7, 8], "sex": [7, 8], "height": [7, 8], "weight": [7, 8], "length_of_stai": [7, 8], "manag": [7, 8], "sever": [7, 8], "alvarado_scor": [7, 8], "paedriatic_appendicitis_scor": [7, 8], "dfn": 7, "convert_dtyp": 7, "python": 7, "sheet": 7, "einf\u00fcgen": 7, "kategorisch": 7, "variablen": 7, "sollten": 7, "per": [7, 8], "hand": 7, "definiert": 7, "wissen": 7, "welcher": 7, "datentyp": 7, "unterschiedlich": 7, "insgesamt": 7, "indik": 7, "kleiner": 7, "gesch\u00e4tzt": 7, "beobachtungen": 7, "kategori": 7, "erkannt": 7, "anpassen": 7, "schwell": 7, "categori": 7, "name": [7, 8], "item": 7, "unique_count": 7, "uniqu": 7, "shape": [7, 8], "total_count": 7, "astyp": 7, "problem": 7, "lenght_of_stai": 7, "valu": 7, "780": 7, "threshold": 7, "m\u00fcsste": 7, "02": 7, "liegen": 7, "diesem": 7, "fall": 7, "fine": 7, "evt": 7, "gestaffelt": 7, "ab": [7, 9], "100": 7, "zahl": 7, "udn": 7, "50": 7, "In": 7, "final": 7, "step": 7, "we": 7, "set": 7, "have": 7, "same": 7, "structur": 7, "number": 7, "your": 7, "origin": 7, "freuenci": 7, "preserv": 7, "albeit": 7, "introduct": 7, "some": 7, "nois": 7, "found": 7, "larg": 7, "similar": 7, "g": 7, "frequenc": 7, "new": 7, "append": 7, "its": 7, "save": 7, "directori": 7, "select_dtyp": 7, "new_col": 7, "seri": 7, "replac": 7, "nan": [7, 8], "nameerror": 7, "traceback": 7, "most": 7, "recent": 7, "call": 7, "last": 7, "cell": 7, "line": 7, "defin": 7, "creat": 7, "synth": 7, "between": 7, "real": [7, 8], "remain": 7, "var": 7, "length": 7, "detect": 7, "recreat": 7, "adress": 7, "faker": 7, "adresslin": 7, "k\u00fcnstlich": [8, 10], "erzeugt": 8, "ihrer": 8, "gesamtheit": 8, "statistischen": [8, 10], "datenstichprob": 8, "gemessenen": 8, "nachbildet": 8, "modellierung": 8, "approxim": 8, "erfordern": 8, "model": 8, "produziert": 8, "aussehen": 8, "bedeutung": 8, "tragen": 8, "datenoperationen": 8, "ergebniss": 8, "liefern": 8, "keiner": 8, "realen": 8, "person": 8, "zugeordnet": 8, "perfekt": 8, "proxi": 8, "urspr\u00fcnglich": 8, "kann": [8, 11], "statt": 8, "eingesetzt": 8, "sdv": 8, "metadata": 8, "singletablemetadata": 8, "lite": 8, "singletablepreset": 8, "head": 8, "7898": 8, "68": 8, "femal": 8, "148": 8, "37": 8, "conserv": 8, "uncompl": 8, "6952": 8, "31": 8, "male": 8, "147": 8, "69": 8, "7909": 8, "23": 8, "163": 8, "62": 8, "8163": 8, "165": 8, "56": 8, "1933": 8, "08": 8, "45": 8, "080000": 8, "900000": 8, "205": 8, "9035": 8, "350000": 8, "600000": 8, "240": 8, "9700": 8, "267625": 8, "335466": 8, "detect_from_datafram": 8, "synthes": 8, "fast_ml": 8, "fit": 8, "med_data_synth": 8, "sampl": 8, "num_row": 8, "7109": 8, "236162": 8, "21": 8, "136330": 8, "176": 8, "644801": 8, "63": 8, "284976": 8, "942523": 8, "087310": 8, "049639": 8, "5765": 8, "050711": 8, "566346": 8, "103": 8, "906050": 8, "438980": 8, "257973": 8, "905003": 8, "881729": 8, "2402": 8, "727097": 8, "196794": 8, "138": 8, "056399": 8, "037305": 8, "902269": 8, "000000": 8, "364234": 8, "8480": 8, "964649": 8, "336835": 8, "143": 8, "513145": 8, "516364": 8, "043838": 8, "704921": 8, "327257": 8, "3978": 8, "201880": 8, "236390": 8, "144": 8, "011407": 8, "35": 8, "515245": 8, "774677": 8, "770871": 8, "f\u00e4llt": 8, "euch": 8, "spoiler": 8, "numerischen": 8, "zum": [8, 9, 10], "beispiel": [1, 8, 9], "mehr": 8, "nachkommastellen": 8, "originalen": 8, "sg": [], "lunch": [], "talk": [], "teil": [], "wo": [], "einsatz": 10, "sinnvollsten": [], "welch": [], "datenschutzbedenken": [], "m\u00f6glichkeiten": [], "mitigieren": [], "fallen": [], "besonderheiten": 9, "praxisbeispiel": 9, "qualit\u00e4tsbewertung": 9, "datenauswertung": 9, "ki": [9, 10], "acknowledg": 9, "praxi": 10, "dummi": 10, "erm\u00f6glichen": 10, "trotz": 10, "datenschutzvorgaben": 10, "versuchsdaten": 10, "softwaretest": 10, "datenbeispiel": 10, "konzepten": 10, "forschungsvorhaben": 10, "produktvorstellungen": 10, "arbeitseb": 10, "projekt": 10, "administr": 10, "uk": 10, "econom": 10, "social": 10, "council": 10, "kritisch": 10, "betrachtung": 10, "datenqualit\u00e4t": 10, "verst\u00e4ndni": 10, "auswirkungen": 10, "vorschl\u00e4g": 10, "ausweitung": 10, "echter": 10, "produkt": 10, "colab": 10, "notebook": 10, "erweitern": 10, "unvollst\u00e4ndig": 10, "erg\u00e4nzen": 10, "bia": 10, "aufheben": 10, "googl": 10, "cloud": 10, "platform": 10, "gretel": 10, "integr": 10, "dienstleistung": 10, "mlop": 10, "plattform": 10, "vertex": 10, "ai": 10, "verbesserung": 10, "train": 10, "ml": 10, "modellen": 10, "sensibl": 10, "internen": 10, "verarbeitung": 10, "scienc": 10, "team": 10, "forschungszwecken": 10, "bereitstellen": 10, "intelligenz": 10, "forschungsdatenzentrum": 10, "bfarm": 10, "erforschung": 10, "anonymisierungsm\u00f6glichkeiten": 10, "readi": 10, "fdz": 10, "vorhaben": 10, "gesundheitsdaten": 10, "erzeugen": 10, "jedoch": 10, "gleichen": 10, "projektbeschreibung": 10, "insbesonder": 11, "forschungsdaten": [], "qualit\u00e4t": 11, "u": 11, "folgenden": 11, "metriken": 11, "gemessen": 11, "\u00fcbersicht": 11, "stammt": 11, "wurd": 11, "eigen": 11, "angaben": 11, "erg\u00e4nzt": 11, "qualit\u00e4tsmetriken": 11, "baum": 11, "englisch": 11, "forschung": 11, "architektur": 1, "neuronalen": 1, "netz": 1, "zufallszahlen": 1, "http": 1, "datacebo": 1, "com": 1, "blog": 1, "interpret": 1, "ctgan": 1, "progress": [1, 9], "wiki": 9, "deren": 9, "potenziellen": 9, "zielt": 9, "praxisnah": 9, "anwendungsorientiert": 9, "einf\u00fchrung": 9, "thema": 9, "dient": 9, "hoffentlich": 9, "allen": 9, "neugierig": 9, "tats\u00e4chlich": 9, "anwendungen": 9, "wollen": 9, "ganz": 9, "\u00fcberzeugt": 9, "sinnvol": 9, "skeptisch": 9, "statistisch": 9, "nachbildbar": 9, "zweifel": 9, "anleitung": 9, "selbermachen": 9, "suchen": 9, "arbeiitsvers": 9}, "objects": {}, "objtypes": {}, "objnames": {}, "titleterms": {"acknowledg": 0, "refer": 0, "besonderheiten": 1, "welch": 1, "arten": 1, "von": [1, 2, 7, 10], "synthetischen": 1, "daten": [1, 2, 5, 6, 7, 8, 9], "gibt": 1, "es": 1, "wa": 1, "macht": 1, "den": 1, "einsatz": 1, "spannend": 1, "datenauswertung": 2, "warum": 2, "brauchen": 2, "wir": 2, "beispiel": [2, 5], "abfrag": [2, 5], "h\u00e4ufigkeiten": 2, "hinwei": 2, "datenschutz": 3, "begriff": 3, "ziel": 3, "differenti": [4, 5], "privaci": [4, 5], "mathematisch": 4, "definit": [4, 8], "intuit": 4, "mit": [5, 6], "und": [5, 10], "synthetisch": [5, 6, 8, 9], "datenschutzkonform": 5, "ki": 6, "bestehend": 6, "regulierungen": 6, "forschungsergebniss": 6, "platz": 6, "1": 6, "5": 6, "generierung": 7, "dummi": 7, "write": 7, "test": 7, "function": 7, "see": 7, "sa": 7, "stata": 7, "file": 7, "run": 7, "properli": 7, "der": [9, 10], "verwaltung": [9, 10], "ein": 8, "datenbeispiel": 8, "teaser": [], "diskuss": [], "praxisbeispiel": 10, "struktur": 10, "aufbau": 10, "datens\u00e4tzen": 10, "veranschaulichen": 10, "echt": 10, "datengrundlagen": 10, "aufwerten": 10, "sch\u00fctzenswert": 10, "teilen": 10, "qualit\u00e4tsbewertung": 11}, "envversion": {"sphinx.domains.c": 2, "sphinx.domains.changeset": 1, "sphinx.domains.citation": 1, "sphinx.domains.cpp": 6, "sphinx.domains.index": 1, "sphinx.domains.javascript": 2, "sphinx.domains.math": 2, "sphinx.domains.python": 3, "sphinx.domains.rst": 2, "sphinx.domains.std": 2, "sphinx.ext.intersphinx": 1, "sphinxcontrib.bibtex": 9, "sphinx": 56}})
\ No newline at end of file